the rise of europe: atlantic trade, institutional change...

The Rise of Europe Atlantic Trade Institutional Changeand Economic Growth

By DARON ACEMOGLU SIMON JOHNSON AND JAMES ROBINSON

The rise of Western Europe after 1500 is due largely to growth in countries withaccess to the Atlantic Ocean and with substantial trade with the New World Africaand Asia via the Atlantic This trade and the associated colonialism affected Europenot only directly but also indirectly by inducing institutional change Whereldquoinitialrdquo political institutions (those established before 1500) placed significantchecks on the monarchy the growth of Atlantic trade strengthened merchant groupsby constraining the power of the monarchy and helped merchants obtain changesin institutions to protect property rights These changes were central to subsequenteconomic growth (JEL F10 N13 O10 P10)

The world we live in was shaped by the rapideconomic growth that took place in nineteenth-century Western Europe The origins of thisgrowth and the associated Industrial Revolutionare generally considered to lie in the economicpolitical and social development of WesternEurope over the preceding centuries In factbetween 1500 and 1800 Western Europe expe-rienced a historically unprecedented period ofsustained growth perhaps the ldquoFirst Great Di-vergencerdquo (ie the first major sustained diver-gence in income per capita across differentregions of the world) making this area substan-tially richer than Asia and Eastern Europe

There is little agreement however on why thisgrowth took place in Western Europe and whyit started in the sixteenth century

This paper establishes the patterns of eco-nomic growth in Western Europe during thisera develops a hypothesis on the origins of therise of (Western) Europe and provides historicaland econometric evidence supporting some ofthe implications of this hypothesis

We document that the differential growth ofWestern Europe during the sixteenth seven-teenth eighteenth and early nineteenth centu-ries is almost entirely accounted for by thegrowth of nations with access to the AtlanticOcean and of Atlantic traders Throughout thepaper the term Atlantic trader refers to BritainFrance the Netherlands Portugal and Spainthe nations most directly involved in trade andcolonialism in the New World and Asia Atlan-tic trade in turn means trade with the NewWorld as well as trade with Asia via the Atlan-tic and includes colonialism- and slavery-related activities1 The differential growth ofAtlantic traders suggests a close link betweenAtlantic trade and the First Great Divergence Infact it appears that the rise of Europe between1500 and 1850 is largely the rise of Atlantic

Acemoglu Department of Economics Massachusetts In-stitute of Technology E52-371 Cambridge MA 02142(e-mail daronmitedu) Johnson Sloan School of Manage-ment Massachusetts Institute of Technology 50 MemorialDrive Cambridge MA 02139 (e-mail sjohnsonmitedu)Robinson Department of Government Harvard University1875 Cambridge Street Cambridge MA 02138 (e-mailjrobinsongovharvardedu) We thank Thomas Becker andRui Pedro Esteves for outstanding research assistance and JoshAngrist Abhijit Banerjee Dora Costa Jan de Vries StanleyEngerman Philip Hoffman Peter Lindert Sebastian MazzucaJoel Mokyr Larry Neal Steve Pincus Christina Romer DavidRomer Andrei Shleifer Alan Taylor Hans-Joachim Voth twoanonymous referees and seminar participants at the Universityof California Berkeley the Canadian Institute of AdvancedResearch University of Chicago Graduate School of Businessand Department of Political Science George Mason Univer-sity Harvard Business School the Harvard Economic HistorySeminar the London School of Economics MIT the NationalBureau of Economic Research economic history inequalityand economic growth groups New York University andPrinceton University for comments and suggestions

1 Atlantic trade opportunities became available only dur-ing the late fifteenth century thanks to the discovery of theNew World and the passage to Asia around the Cape ofGood Hope These discoveries resulted from a series ofinnovations in ship technology primarily pioneered by thePortuguese that changed the rigging and hull design ofships and developed knowledge of oceanic navigation

546

Europe and is quite different in nature from theEuropean growth that took place before 1500

Not all societies with access to the Atlanticshow the same pattern of growth however Thedata suggest an important interaction betweenmedieval political institutions and access to theAtlantic the more rapid economic growth tookplace in societies with relatively nonabsolutistinitial institutions most notably in Britain andthe Netherlands In contrast countries where themonarchy was highly absolutist such as Spainand Portugal experienced only limited growthin the subsequent centuries while areas lacking

easy access to the Atlantic even such nonabso-lutist states as Venice and Genoa did not expe-rience any direct or indirect benefits fromAtlantic trade

Figures 1 and 2 illustrate the central thesisof this paper Figure 1 panel A shows thaturbanization in Western Europe grew sig-nificantly faster than in Eastern Europe after15002 Figure 1 panel B shows that these

2 For the purposes of this paper Western Europe is takento be all the countries west of the Elbe ie Austria

FIGURE 1A WESTERN EUROPE EASTERN EUROPE AND ASIA URBANIZATION RATES WEIGHTED BY POPULATION 1300ndash1850

FIGURE 1B ATLANTIC TRADERS WEST EUROPEAN COUNTRIES NOT ATLANTIC TRADERS AND EASTERN EUROPEURBANIZATION RATES WEIGHTED BY POPULATION 1300ndash1850

547VOL 95 NO 3 ACEMOGLU ET AL THE RISE OF EUROPE

differential trends are due in large part tothe growth of Atlantic traders The rest of West-ern Europe had a relatively high average urban-ization rate of 10 percent in 1300 (and 114percent in 1500) but grew at approximately thesame rate as Eastern Europe from 1500 to 1850by a factor of less than 2 to reach 17 percent by

1850 In contrast Atlantic traders started with alower average urbanization rate of 8 percent in1300 (and only 101 percent in 1500) whichalmost tripled in the subsequent 550 years toreach 245 percent in 1850 overtaking averageurbanization in the non-Atlantic parts of West-ern Europe between 1600 and 1700 (see Table 1)Panels A and B in Figure 2 show the samepattern using Angus Maddisonrsquos (2001) esti-mates of GDP per capita While GDP per capitarose by a factor of almost two among Atlantictraders between 1500 and 1820 in the rest ofWestern Europe it grew at approximately thesame rate as in Eastern Europe just under 30percent

The patterns depicted in Figures 1 and 2 do

Belgium Britain Denmark Finland France Germany Ire-land Italy the Netherlands Norway Portugal Spain Swe-den and Switzerland Eastern Europe is all Europeancountries to the east of the Elbe including Russia andexcluding Turkey See Section I A for details on urbaniza-tion and GDP data All averages are weighted by popula-tion using numbers from Colin McEvedy and RichardJones (1978)

FIGURE 2B ATLANTIC TRADERS WEST EUROPEAN COUNTRIES NOT ATLANTIC TRADERS AND EASTERN EUROPE GDP PER

CAPITA WEIGHTED BY POPULATION 1500ndash1870

FIGURE 2A WESTERN EUROPE EASTERN EUROPE AND ASIA GDP PER CAPITA WEIGHTED BY POPULATION 1500ndash1870

548 THE AMERICAN ECONOMIC REVIEW JUNE 2005

not simply reflect the tendency of more suc-cessful nations to engage in Atlantic tradeThere is no differential growth of Atlantictraders before the opening of Atlantic searoutes and below we show similar resultsusing an exogenous measure of access to theAtlanticmdashratio of Atlantic coastline to landareamdashinstead of the distinction between At-lantic traders and nontraders Nor do the re-sults reflect some post-1500 advantage of

coastal nations Atlantic ports grew muchfaster than other European cities while Med-iterranean ports grew at similar rates to inlandcities

This evidence weighs against the most pop-ular theories for the rise of Europe which em-phasize the continuity between pre-1500 andpost-1500 growth and the importance of certaindistinctive European characteristics such asculture religion geography and features of the

TABLE 1mdashDESCRIPTIVE STATISTICS

Wholesample

unweighted

Wholesample

weighted

AtlanticWesternEurope

Non-AtlanticWesternEurope

EasternEurope Asia

Weighted by population

Urbanization in 1300 66 99 80 100 41 110(52) (32) (28) (61) (33) (07)

Urbanization in 1400 76 103 85 121 39 111(95) (36) (24) (100) (15) (05)

Urbanization in 1500 83 106 101 114 40 115(76) (34) (53) (68) (18) (07)

Urbanization in 1600 96 117 136 140 44 120(76) (40) (76) (88) (27) (07)

Urbanization in 1700 107 112 145 131 37 116(85) (41) (68) (81) (22) (07)

Urbanization in 1800 141 103 198 169 70 89(91) (49) (79) (75) (33) (14)

GDP per capita in 1500 62754 6083 72146 85073 50694 5750(1593) (1180) (311) (2171) (782) (354)

GDP per capita in 1600 74073 6305 91631 90822 57829 5768(2256) (1442) (1493) (1673) (1123) (353)

GDP per capita in 1700 86212 6222 107921 98082 6360 5742(3484) (2081) (3214) (1282) (1361) (353)

GDP per capita in 1820 98800 6917 132195 109540 7195 5755(3736) (2645) (3487) (1253) (1749) (457)

Constraint on executive in 1500 167 173 175 199 146(076) (079) (056) (099) (079)




Atlantic coastline-to-area 00057 00014 00118 00026 000 000(00117) (00065) (00181) (00052)

Notes First column is unweighted means other columns are mean values weighted by total population in year indicated fromMcEvedy and Jones (1978) Standard deviation is in parentheses There are 24 European countries in these data AtlanticWestern Europe is England France the Netherlands Portugal and Spain Non-Atlantic Western Europe is Austria BelgiumDenmark Finland Germany Ireland Italy Norway Sweden and Switzerland Eastern Europe is Albania Bulgaria theCzech Republic Greece Hungary Poland Romania Russia and Serbia Asia is India and China Urbanization for Europeis percentage of population living in towns with population of at least 5000 at some time between 800 and 1800 from PaulBairoch et al (1988) for Europe comparable data for Asia are from Bairoch (1998) GDP per capita is from Maddison (2001)Constraint on executive is on a scale of 1 to 7 where a higher score indicates more constraints this is coded using the PolityIV methodology as explained in the text We have not coded constraint on the executive for Asia Atlantic coast-to-areaincludes those parts of Germany Denmark and Norway that are on the North Sea For more detailed definitions and sourcessee Appendix Table 1


European state system3 Instead it is consistentwith theories that emphasize the importance ofprofits made in Atlantic trade colonialism andslavery4 Nevertheless other evidence suggeststhat overseas trade and the associated profitswere not large enough to be directly responsiblefor the process of growth in Europe Stanley LEngerman (1972) and Patrick K OrsquoBrien(1982) demonstrate that the contribution ofprofits from slavery and trade with the rest ofthe world to European capital accumulation wasmodest OrsquoBrien (1982 p 2) writes that trans-oceanic trade ldquo could in no way be classifiedas decisive for economic growth of WesternEuroperdquo Although recent work by Joseph EInikori (2002) estimates larger trade flows thanthose of OrsquoBrien his estimates are not largeenough to suggest that European growth wasdriven solely by the direct impact of Atlantictrade on profits or resources

We advance the hypothesis that West Euro-pean growth during this period resulted in partfrom the indirect effects of international tradeon institutional development Although therewere some improvements in economic institu-tions in the late medieval and early modernperiod rapid economic development did notbegin until the emergence of political institu-tions providing secure property rights to abroader segment of society and allowing freeentry into profitable businesses (Douglass CNorth and Robert P Thomas 1973 North and

Barry R Weingast 1989) The critical politicalinstitutions were those that constrained thepower of the monarchy and allied groups5

Checks on royal power and prerogativesemerged only when groups that favored themthat is commercial interests outside the royalcircle became sufficiently powerful politicallyFrom 1500 and especially from 1600 onwardin countries with nonabsolutist initial institu-tions and easy access to the Atlantic the rise inAtlantic trade enriched and strengthened com-mercial interests outside the royal circle andenabled them to demand and obtain the institu-tional changes necessary for economic growthAlthough profits from Atlantic trade were smallrelative to GDP they were still substantial andmuch larger than previous trading profits Forexample Figure 3 shows that by the end of theseventeenth century the volume of Atlantictrade was much larger than that of long-distanceMediterranean trade (see the Appendix for theconstruction of these series) The recipients ofthese profits became very rich by the standardsof seventeenth- and eighteenth-century Europeand typically politically and socially verypowerful

These changes did not take place in countrieswith highly absolutist institutions such as Spain

3 See eg Max Weber (1905) Eric Jones (1981) JohnA Hall (1985) and David S Landes (1998)

4 Eg Eric E Williams (1944) Andre Gunder Frank(1978) and Immanuel M Wallerstein (1974ndash1980)

5 It is important to note that these new political institu-tions neither protected the rights of all citizens nor weredemocratic They can best be characterized as oligarchicsince they increased the political power of wealthy mer-chants and at least in the British case of the gentry andnascent industrial interests Nevertheless they constituted adistinct improvement over the previous set of institutionswhich placed many fewer checks on the power of themonarchy

FIGURE 3 VOLUME OF ATLANTIC AND MEDITERRANEAN TRADE (VOYAGE EQUIVALENTS PER YEAR) 1300ndash1800


Portugal and to a large extent France where thecrown was able to closely control the expansion oftrade Consequently in these countries it was themonarchy and groups allied with it that were themain beneficiaries of the early profits from Atlan-tic trade and plunder and groups favoring changesin political institutions did not become powerfulenough to induce them Our hypothesis thereforepredicts an important interaction between initialinstitutions and Atlantic trade which is the patternwe find in the data

The major premise presented in this paper isconsistent with the emphasis of a number ofhistorians including among others RalphDavis (1973a) Jan de Vries (1984) PaulBairoch (1988) Fernand Braudel (1992) and deVries and Ad van der Woude (1997) Althoughthis historical literature emphasizes the differ-ential growth of Atlantic ports and Atlantic na-tions to the best of our knowledge there are noother studies documenting the quantitative im-portance of Atlantic traders and Atlantic portsor showing that the differential growth of West-ern Europe is accounted for largely by thegrowth of Atlantic traders

On the theoretical side our hypothesis buildson the notion that institutional change evenwhen socially beneficial will be resisted bysocial groups that stand to lose economic rentsor political power Consequently the process ofinstitutional change involves significant conflictbetween different groupsmdashin the Europeancontext between the monarchy and its alliesversus commercial interests outside the royalcircle6 Our historical account can also beviewed as a marriage between the Marxist thesislinking the rise of the bourgeoisie and the de-velopment of the world economy (eg amongothers Williams 1944 Frank 1978 andWallerstein 1974ndash1980) and the neoclassicalemphasis on the development of political insti-tutions and secure property rights in WesternEurope (eg North and Thomas 1973 Eric LJones 1981 North 1981 J Bradford De Long

and Andrei Shleifer 1993) Distinct from theseapproaches however we offer an explanationbased on the interaction between Atlantic tradeand medieval political institutions of whystrong private property rights emerged in West-ern Europe especially in Britain and the Neth-erlands starting in the sixteenth centuryAlthough some scholars have noted the impor-tant role of overseas merchants in particularinstances of political change during this period(most notably Robert Brenner 2003 andSteven Pincus 2002 in the British case) we arenot aware of a theory along the lines developedin this paper

The paper is organized as follows Section Idocuments the key premise of the paper andshows that the pattern seen in Figures 1 and 2 isrobust Section II develops our hypothesis forthe rise of Europe and the role played by At-lantic trade in this process and provides histor-ical evidence supporting our interpretationSections III and IV provide evidence on someimplications of our hypothesis (Section IIIshows that the evolution of European institu-tions is closely linked to Atlantic trade andSection IV documents an important interactionbetween initial institutions and Atlantic trade inEuropean economic growth) Section V con-cludes The Appendix summarizes the construc-tion of the variables used in the empiricalanalysis and further detail can be found inAcemoglu et al (2002b)

I Atlantic Trade and the Rise of Europe

A Data

We use three data series to measure economicdevelopment First we construct estimates ofurbanization based on the urban populationnumbers of Bairoch et al (1988) This is acomprehensive dataset with information on all2200 European cities that had at some timebetween 800 and 1800 5000 or more inhabit-ants7 We use these data as our measure ofurban population and divide by the population6 See for example North (1981) Mancur Olson (1982)

Per Krusell and Jose-Victor Rios-Rull (1996) Stephen Par-ente and Edward C Prescott (1999) Acemoglu and Robin-son (2000 2002) and Raghuram G Rajan and LuigiZingales (2003) Ronald Rogowski (1989) is particularlynotable in this context since he also emphasizes how tradeaffects political coalitions via its impact on factor pricesalthough he does not focus on how trade might induceinstitutional change by strengthening commercial interests

7 These data begin in 800 and there are estimates forevery 100 years until 1700 then for every 50 years through1850 Bairoch et al (1988) emphasize however that esti-mates before 1300 are rough and less reliable (and they skipthe year 1100 due to lack of information) These data wereused previously by De Long and Shleifer (1993)


estimates of McEvedy and Jones (1978) to cal-culate urbanization (percentage of the popula-tion living in cities with more than 5000inhabitants) We also use estimates of urbaniza-tion rates for Asia from the quantitative andqualitative assessments of Bairoch (1988)Bairoch (1988 ch 1) and de Vries (1976 p164) argue that only areas with high agriculturalproductivity and a developed transportation net-work could support large urban populations Inaddition in Acemoglu et al (2002a) we pre-sented evidence that both in the time series andthe cross section there is a close associationbetween urbanization and income per capitabefore as well as after industrialization Wetherefore take urbanization as a proxy for GDPper capita

Second we use estimates of GDP per capitafrom Maddison (2001) These estimates start in1500 and are available for 1600 1700 1820and then more frequently Note that these esti-mates are no more than educated guesses espe-cially before 1820 We therefore think of theseGDP data as a check on our results using ur-banization data

Third we use the European cityndashlevel datafrom Bairoch et al (1988) to investigate whichurban centers were driving demographic andeconomic growth and also to contrast thegrowth of Atlantic ports to other ports and toinland cities

Table 1 gives the estimates of urbanizationand income per capita at various dates The firstcolumn is for the whole sample and is un-weighted The second column is weighted bypopulation in the corresponding year giving abetter sense of the aggregate changes The re-maining columns give weighted means for At-lantic traders (Britain France the NetherlandsPortugal and Spain) for West European coun-tries that were not Atlantic traders (AustriaBelgium Denmark Finland Germany IrelandItaly Norway Sweden and Switzerland) forEast European countries and for the Asiancountries in our sample8 These numbers con-

firm the patterns shown in Figures 1 and 2 Inthe regression analysis we will report bothweighted and unweighted results The bottomthird of the table also shows the evolution of ourmeasure of institutions constraint on the exec-utive which we will be described in greaterdetail and used in Section III

B Economic Growth in Europe

Figures 1A and 1B show the evolution ofurbanization rates in Western and Eastern Eu-rope and contrast the behavior of Atlantic trad-ers versus non-Atlantic traders We first look atAtlantic traders since the main beneficiariesfrom the Atlantic should be those countries thatengaged in Atlantic trade and colonialismHowever whether or not a country is an Atlan-tic trader is clearly endogenous ie it is theoutcome of some political or economic processFor this reason we also present results using ameasure of access to the Atlantic which is acountry-level geographic characteristic

We can test the idea that West Europeangrowth after 1500 was due primarily to growthin countries involved in Atlantic trade or with ahigh potential for Atlantic trade by estimatingthe following regression equation

(1) ujt dt j t1600

t WEj dt

t1500

t PATj dt Xjt jt

where ujt is urbanization in country j at time tWEj is a dummy indicating whether the countryis in Western Europe the dtrsquos denote year ef-fects the jrsquos denote country effects Xjt is avector of other covariates and jt is a distur-bance term In addition PATj our measure ofthe potential for Atlantic trade is a dummy forAtlantic trader (Britain France the Nether-lands Portugal and Spain) or alternatively theAtlantic coastline-to-area ratio (in both cases atime-invariant characteristic of the country)The trsquos the coefficients on the potential for

8 We take current countries as the unit of observationAlthough these do not always correspond to the independentpolities of the time this discrepancy should not bias ourempirical inference For example if we had data on eachItalian city-state their average would show the same patternas our single Italy observation (presuming that our data forthe aggregate of Italy are accurate) but because of the larger

number of observations the standard errors would besmaller The analysis of city-level growth in Section I B isinformative on differential growth across historical politicalboundaries


Atlantic trade and the post-1500 time dummiesare the main parameters of interest Since ourfocus is on the rise of Western Europe as awhole our basic regressions are weighted bypopulation in each year but we also reportunweighted regressions for completeness

Columns 1 and 2 of Table 2 include only theinteraction terms between the Western Europedummy and dates from 1600 yent1600 t WEj dt which capture the differential growth ofWest European countries relative to Eastern Eu-rope The top row reports the p-value from theF-test of the joint significance of these interac-tions Column 1 includes data only for 1300ndash1850 while column 2 extends the sample backto 1000 Consistent with Figure 1A both spec-ifications show significantly faster growth in

Western Europe than in Eastern Europe Forexample the point estimates (not shown in thetable to save space) indicate that in the specifi-cation of column 1 West European urbaniza-tion grew by 69 percentage points relative toEast European urbanization between 1500 and1850

Column 3 allows differential growth forcountries engaged in Atlantic trade by includ-ing the term yent1500 t PATj dt We include1500 as a ldquospecification checkrdquo on the timing ofthe effects We start with PATj as a dummy forAtlantic trader Significant positive estimates oftrsquos imply that Atlantic traders grew starting inthe period between 1500 and 1600 The esti-mates confirm the pattern seen in Figure 1B andshow large effects from the interaction between

TABLE 2mdashATLANTIC TRADE AND URBANIZATION

Dependent variable is country-level urbanization

Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850unweighted

Panel1300ndash1850with Asia

Panel1300ndash1850

without BritainPanel

1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

Potential for Atlantic trade is measured by

Atlantic trader dummy Atlantic coastline-to-area

Panel A Flexible specification

p-value for Western Europe year dummies 1600ndash1850

[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]

Potential for Atlantic trade 1500

0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)

R-squared 087 085 089 087 082 084 093 094 092 083Number of observations 192 240 192 240 192 208 184 192 240 192

Panel B Structured specification


[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]

Potential for Atlantic trade 0011 00083 0016 0011 0005 075 065 062 volume of Atlantictrade

(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


Notes Standard errors are in parentheses Panel regressions with full set of country and year dummies regressions are weighted unless otherwise stated Weightedregressions use total population in each year as weights from McEvedy and Jones (1978) Dependent variable is level of urbanization (percentage of population livingin towns that had at least 5000 population at some point between 800 and 1800) in each country in each year Urbanization in Europe is from Bairoch et al (1988)and urbanization in Asia is from Bairoch (1998) We report results with two different measures of potential for Atlantic trade a dummy for whether a country wasan Atlantic trader (one for Britain the Netherlands France Spain and Portugal zero for all others) in columns 3 4 5 6 and 7 and the ratio of Atlantic coastlineto area for the Atlantic trader countries plus Belgium Denmark Germany Ireland and Norway (columns 8 9 and 10) Column 6 includes the available data on Asia(just for India and China) and column 7 drops the data for Britain Volume of Atlantic Trade is the log average number of voyages per year For more detailed datadefinitions and sources see Appendix Table 1


the Atlantic trader dummy and dates after 1600These effects become statistically significant af-ter 1750 in columns 8ndash10 the effects are sta-tistically significant starting in 1700 Forexample the estimate for 1850 1850 0085implies that urbanization among Atlantic trad-ers grew by approximately 85 percentagepoints more than in other Western and EasternEuropean nations Notice also that the estimateof 1500 in this column which measures thedifferential growth of Atlantic traders between1300ndash1400 and 1500 is insignificant and smallThis is reassuring since Atlantic trade was verylimited before 1500 this finding shows thatthere is no differential growth for Atlantic trad-ers before Atlantic trade actually becameimportant9

Consistent with the patterns shown in Fig-ure 1B the inclusion of the Atlantic tradeinteractions explains almost the entire differ-ential growth of West European nations rela-tive to Eastern Europe The yent1600 t WEj dt terms are no longer statistically significantand the point estimates (not shown in thetable) imply that West European urbanizationgrew only by 29 percentage points relative toEastern Europe between 1300 ndash1500 and1850 as opposed to 69 percentage points incolumn 1

Columns 4 and 5 show that the results aresimilar for the 1000ndash1850 period and whenobservations are not weighted by population10

Column 6 includes Asian countries This haslittle effect on the estimates of the differentialgrowth of Atlantic traders but now West Euro-

pean countries are growing faster relative to thecontrol group which includes Asian countries(see Figure 1A) Finally column 7 excludesBritain from the sample and shows that theresults do not simply reflect British growth Theestimates in column 7 are about half the size ofthose in the other columns but they show thesame pattern

An important concern with the results re-ported so far is endogeneity Being an Atlantictrader is an ex post outcome and perhaps onlycountries with high growth potentialmdashor thosethat were going to grow anywaymdashengaged insubstantial Atlantic trade and colonial activityBelgium Ireland Denmark Germany and Nor-way also had access to the Atlantic either di-rectly or via the North Sea but they did not takea major part in long-distance oceanic trade Incolumns 8 9 and 10 we use a geographicmeasure of potential access to the AtlanticAtlantic coastline-to-area ratio as our time-invariant PATj variable which gives positiveAtlantic trade potential to all these countries11

This measure allows Atlantic trade to play amore important role in the growth of countrieswith more Atlantic coastline relative to theirland area12

9 Although the analysis above does not count Denmarkand Sweden as Atlantic traders Sweden had a small colonyon the Delaware river 1637ndash1681 and Denmark controlledseveral small Caribbean islands (now the US Virgin Is-lands) To check the robustness of our results we alsoexperimented with a more inclusive definition of Atlantictrader that includes Denmark and Sweden with results verysimilar to those reported in column 3 The p-value forWestern Europe year interactions increases to [051]while the pattern of coefficients on potential for Atlantictrade year dummies is largely unchanged the interactionsbefore 1700 are insignificant then 0035 (se 0022) in1700 0035 (se 0021) in 1750 0046 (se 002) in1800 and 008 (se 002) in 1850

10 In column 4 the interaction between the West Euro-pean dummy and the post-1500 dates is significant at the10-percent level which reflects the lower level of EastEuropean urbanization in the base period which is now1000ndash1400

11 Information on the length of coastline and the landarea of particular countries is taken from IntegratedCoastline Management (httpicmnoaagovcountryICM-prohtml) which reports a standardized measureWe use only Atlantic coastline ie omitting coastlines inthe Mediterranean the Baltic and the Arctic Details areprovided in the Appendix of Acemoglu et al (2002b) Itis important to exclude the Baltic coastlines of Denmarkand Germany from our measure since significant Baltictrade predated the rise of Atlantic trade and economicgrowth driven by Baltic trade could be an alternativeexplanation for the patterns we observe In any case ourresults are generally robust to including the Baltic or theArctic coastlines For example we obtain very similarresults to those reported in Tables 2 and 3 when weinclude the west coastline of Sweden or when we includethe entire Norwegian coastline on the Arctic and theentire German coastline on the Baltic Our results are alsogenerally similar when we include all the coastline ofSweden Germany Norway and the entire Baltic coast-line of Denmark but the size of the coastline-to-areatimes year interactions are smaller than in our baselineand Western Europe times year interactions become sig-nificant

12 Alternatively we could use the Atlantic coastline-to-area measure as an instrument for the Atlantic traderdummy The results we report can be thought of as thereduced form for this IV strategy (a univariate regression ofthe Atlantic trader dummy on the coastline-to-area measure


The results using the coastline-to-area mea-sure for PATj are similar to those using theAtlantic trader dummy Most notably the dif-ferential growth related to the Atlantic nowcaptured by interactions with the Atlanticcoastline-to-area ratio is still strong the pointestimates for the rsquos are significant starting in1700 and quantitatively large For example thecoefficient 1850 505 indicates approxi-mately 65 percentage points more urbanizationgrowth in the Netherlands than in Italy between1300ndash1400 and 1850 (the Atlantic coastline-to-area ratio for the Netherlands is 0013 and forItaly it is 0) This explains over half of thedifferential 12-percentage-point actual urban-ization growth between Italy and the Nether-lands between these two dates Otherspecifications using the Atlantic coastline-to-area measure in columns 9 and 10 give similarresults

Equation (1) allows for an arbitrary pattern ofdifferential growth in Atlantic traders Insteadwe might expect the differential growth of At-lantic traders to be related to the volume ofAtlantic trade For this reason in panel B wereport results from estimating a structuredmodel of the form

(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt

where ATt denotes our estimate of the aggregatevolume of Atlantic trade shown in Figure 3The construction of this variable is explainedbriefly in the Appendix and further details androbustness results can be found in Acemoglu etal (2002b)

Note that the model in equation (2) is morerestrictive than that in (1) since we are forcingthe pattern of trsquos in (1) to be the same as that

of ln ATt In all columns the estimate of thecoefficient on the interaction term between thelog volume of Atlantic trade and potential forAtlantic trade at the country level is highlysignificant while the interaction terms betweenWestern Europe and dates from 1600 onwardare again insignificant Notably the R2 of thismore restrictive regression is close to the R2 ofthe flexible specifications reported in panel AThese results suggest that the significant inter-action between potential for Atlantic trade anddates after 1600 is due to the importance ofAtlantic trade not some other parallel process

Table 3 which has the same structure asTable 2 provides regression evidence using logGDP per capita as the dependent variable Mad-dison (2001) reports estimates of GDP per cap-ita for 1500 1600 1700 1820 and 1870 Wetake 1500 as the base year and add interactionsbetween our measure of potential for Atlantictrade PATj and the dates from 1600 on tocapture the importance of Atlantic trade forthe country (so we can no longer test for pre-existing trends using the interaction betweenPATj and 1500) Output numbers for 1870 arealready heavily influenced by differential indus-trialization experiences of various countries soour baseline specification stops in 1820 Forcompleteness we also report regressions thatextend the sample to 1870

Parallel to our results in Table 2 West Euro-pean countries grow faster after 1500 althoughthis pattern is somewhat less pronounced espe-cially when we limit the sample to 1500ndash1820The interactions between the Atlantic traderdummy and the dates after 1600 are typicallysignificant starting either in 1600 or 1700 andquantitatively large For example the estimate of1820 027 in column 3 indicates that Atlantictraders grew on average 31 percent (027 logpoints) more than non-Atlantic trader West Euro-pean nations between 1500 and 1820 Columns 4to 7 report similar results to those in Table 2 Thepattern is the same when the sample is extended to1870 with unweighted regressions when Britainis excluded from the sample and when Asiancountries are included Columns 8 to 10 reportsimilar results using the Atlantic coastline-to-areameasure

Panel B of Table 3 reports structured modelssimilar to (2) where we include the interactionterm PATj ln ATt instead of the full set ofpost-1500 interactions between PATj and time

in our sample has an R2 of 030) Nevertheless we prefer thespecification in the text since it is plausible that evenconditional on being an Atlantic trader a country withgreater Atlantic coastline will trade and grow more thananother with less coastline making such an IV procedureinvalid In fact a comparison of columns 3ndash7 with columns8ndash10 shows that the fit of the models with the Atlanticcoastline-to-area ratio is marginally better than those withthe Atlantic trader dummy because the former measuregives greater potential for trade to Britain and the Nether-lands which have relatively high coastline-to-area ratios


dummies This more structured specificationagain shows that the differential growth ofWestern Europe from 1600 is closely linked tothe extension of Atlantic trader

Overall both Table 2 and Table 3 show animportant role for Atlantic trade in West Euro-pean growth When the effect of Atlantic tradeis not taken into account the estimates of trsquosare significant positive and large Western Eu-rope is growing faster than Eastern Europe andAsia Once Atlantic trade interactions are in-cluded trsquos are typically no longer significantwhile the effect of Atlantic trade is very strongFurthermore the estimates show no evidence ofdifferential growth by Atlantic traders beforethe age of Atlantic trade

C Other Determinants of Economic Performance

To check the robustness of our resultsTable 4 adds a number of covariates to our basic

regressions The overall patterns are not af-fected To save space Table 4 reports only thestructured specifications of equation (2)

Weber (1905) and Landes (1998) argue thatreligion is an important determinant of eco-nomic and social development To assess theimportance of religion we allow Protestantcountries to grow at rates different from non-Protestant countries by interacting a dummy forbeing a majority Protestant country in 1600with year dummies starting in 160013 The p-values from the joint significance test reportedin columns 1 of panels A and C show that whenthe dependent variable is the urbanization ratethese interactions are either insignificant or onlymarginally significant In contrast when thedependent variable is log GDP per capita and

13 See the Appendix for the construction of the variablesused in this subsection

TABLE 3mdashATLANTIC TRADE AND GDP PER CAPITA

Dependent variable is country-level log GDP per capita

Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)




p-value for Western Europe year dummies 1600ndash1820 orndash1870

[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]

Potential for Atlantic trade volume of Atlantic trade

0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)


Notes Standard errors are in parentheses Panel regressions with full set of country and year dummies regressions are weighted unless otherwise stated Weightedregressions use total population in each year as weights from McEvedy and Jones (1978) Dependent variable is log GDP per capita from Maddison (2001) We reportresults with two different measures of potential for Atlantic trade a dummy for whether a country was an Atlantic trader (one for Britain the Netherlands FranceSpain and Portugal zero for all others) in columns 3 4 5 6 and 7 and the ratio of Atlantic coastline to area for the Atlantic trader countries plus Belgium DenmarkGermany Ireland and Norway (columns 8 9 and 10) Column 6 includes the available data on Asia (just for India and China) and column 7 drops the data for BritainVolume of Atlantic trade is the log average number of voyages per year For more detailed data definitions and sources see Appendix Table 1


we use the Atlantic trader dummy for our po-tential Atlantic trade measure (panel B) there isa significant effect from these religion timesyear interactions Nevertheless this has littleimpact on the pattern of differential growthbetween Western and Eastern Europe or be-tween Atlantic and non-Atlantic traders More-over the quantitative effects of Protestantismon economic growth are smaller than those ofAtlantic trade14

Many social scientists view war-making asan important factor in the process of statebuilding and subsequent economic develop-ment (eg Otto Hintze 1975 Paul Kennedy1987 Charles Tilly 1990) Incidence of warsmight also proxy for the importance of inter-state competition which many historians in-cluding Jones (1981) and Hall (1985) haveemphasized To assess the importance ofwars in columns 2 and 6 we include a vari-able which is the average number of years atwar during the previous period (a century or

14 The point estimates (not reported) imply that Protes-tant countries experienced 45 percentage points greaterurbanization growth between 1500 and 1850 and 30 percentmore GDP growth between 1500 and 1820 The correspond-ing numbers for Atlantic traders in the flexible specifica-

tions including the Protestant dummy interacted with datesfrom 1600 are 84 percentage points more urbanization and41 percent more GDP growth

TABLE 4mdashROBUSTNESS CHECKS

Panel1300ndash1850controllingfor religion

Panel1300 to 1850

controllingfor wars

Panel1300 to 1850

controllingfor Roman

heritage

Panel1300 to 1850

controllingfor latitude


Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)

Using Atlantic trader dummy measure of potential for Atlantic trade

Panel A Dependent variable is level of urbanization Panel B Dependent variable is log GDP per capita


[067] [042] [049] [009] [024] [091] [015] [085]

Atlantic trader dummy volumeof Atlantic trade

0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)

p-value for Protestant year [007] [000]Wars per year in preceding

century00006 0075(0008) (0029)

p-value for Roman heritage year

[089] [000]

p-value for latitude year [011] [000]R-squared 089 089 089 089 097 095 097 097Number of observations 192 176 192 192 96 88 96 96

Using Atlantic coastline-to-area measure of potential for Atlantic trade

Panel C Dependent variable is level of urbanization Panel D Dependent variable is log GDP per capita


[019] [023] [039] [009] [099] [098] [071] [081]

Coastline-to-area volume ofAtlantic trade

079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033

(0007) (0026)p-value for Roman heritage

year[077] [032]


Notes Standard errors are in parentheses Weighted panel regressions with full set of country and year dummies Weights are total population of countryin each year from McEvedy and Jones (1978) Dependent variable in panels A and C is level of urbanization (percent of population living in towns withmore than 5000 population) Urbanization in Europe is from Bairoch et al (1988) Dependent variable in panels B and D is log GDP per capita from Maddison(2001) Panels A and B use the Atlantic trader dummy as the measure of potential for Atlantic trade (one for Britain France Spain Portugal andthe Netherlands zero for all others) Panels C and D use the ratio of Atlantic coastline to area Volume of Atlantic Trade is the log average number ofvoyages per year Protestant is a dummy for whether country was majority Protestant in 1600 Protestant year is the Protestant dummy interacted with yeardummies for 1600 and after Wars per year are in preceding century through 1700 1700 ndash1750 for 1750 1750 ndash1800 for 1800 and 1800 ndash1850 for 1850 Romanheritage is dummy for whether country was in the Roman Empire this is interacted with year dummies for 1600 and after Latitude is distance from the equatorfor capital city of this country today this is interacted with year dummies for 1600 and after For more detailed data definitions and sources see AppendixTable 1


half-century) We find that this variable itself isinsignificant in the urbanization regressions andhas no effect on the patterns documented so far15

A popular view sees the roots of Europeangrowth in the Roman Empire (eg PerryAnderson 1974 Jones 1981 Landes 1998)and perhaps in the culture of Ancient GreeceTo investigate whether Roman heritage is im-portant for the rise of Europe we created adummy that indicates whether a country waspart of the Roman Empire We then interactedthis variable with dates from 1600 onward tosee whether there is differential growth de-pending on the extent of Roman heritage (col-umns 3 and 7) These interactions aretypically insignificant and do not affect thepatterns reported in the previous tables Theonly exception is when we use log GDP percapita as the dependent variable and the At-lantic trader dummy for PATj But in thiscase the results indicate that countries withRoman heritage grew more rapidly between1400 and 1600 and significantly more slowlythereafter

Finally in columns 4 and 8 we add inter-actions between distance from the equator(the absolute value of the latitude of the na-tionrsquos capital) and dates from 1600 to seewhether the move of economic activity awayfrom Southern toward Northern Europe canexplain the rise of Atlantic nations Onceagain the addition of these variables does notaffect the importance of Atlantic trade andthe latitude interactions are typically insignif-icant (except in panel B where the pointestimates have the wrong sign)

D Urban Expansion and Atlantic Ports

We next turn to an analysis of data on thepopulation of individual cities compiled by

Bairoch et al (1988) Figure 4A shows that theurban expansion of Western Europe was drivenby cities that were Atlantic ports Table 5 con-firms this pattern with regression analysis Itestimates models similar to (1) with the log ofcity-level urban population as the dependentvariable The key right-hand side variable isthe interaction between a dummy indicatingwhether the city is an Atlantic port (or in ouralternative specification whether it is a po-tential Atlantic port) denoted by APi anddummies from 150016 The sample for all re-gressions in Table 5 is the balanced panel ofcities for which we have observations in eachdate17

In column 1 APi is a dummy for Atlanticport and observations are weighted by cur-rent population in each year The interactionsbetween the Atlantic port dummy and datesafter 1600 the APi dt terms are statisticallyand economically significant and positive Forexample the coefficient of 079 implies thatAtlantic ports grew approximately 120 per-cent (079 log points) relative to other citiesbetween 1300 ndash1400 and 1800 Notably thereappears to be no differential growth of Atlan-tic ports before 1600 once again supportingthe notion that the growth of these ports isrelated to the emergence of trading and colo-nial opportunities via the Atlantic In the bot-tom panel we report results from a structuredspecification similar to equation (2) Onceagain the coefficient on the interaction termbetween the volume of Atlantic trade and the

15 As an alternative exercise more favorable to the warhypothesis we also controlled for the average number ofyears at war that ended in victory during the previous 50 or100 years To the extent that rich nations are more likely tosucceed in war the coefficient on this variable will bebiased upward The inclusion of this variable has remark-ably little effect on our estimates of the interaction betweenaccess to the Atlantic (or Atlantic trader) and the post-1500years (or the volume of Atlantic trade) and this war variableitself is insignificant when the dependent variable is theurbanization rate and marginally significant with log GDPper capita

16 See the Appendix of Acemoglu et al (2002b) for thelist of Atlantic ports in our panel In Figures 4 and 5 we usethe definition of actual Atlantic port In the regressionanalysis we also report results with a dummy for potentialAtlantic port The distinction between Atlantic port andpotential Atlantic port parallels our use of Atlantic traderdummy and the coastline-to-area measure of potential forAtlantic trade in Tables 2 3 and 4

17 The focus on a balanced panel of cities avoidsproblems of composition bias which would result fromthe fact that cities enter the dataset only once they exceeda certain threshold (typically 5000 people) For exampleif an area is growing rapidly the population of thesmaller cities in this area will also grow and exceed therelevant threshold but the addition of cities with popu-lation around 5000 may reduce the average population ofthe cities in this area Nevertheless in practice this biasdoes not seem to be important and in Acemoglu et al(2002b) we report similar results using a larger unbal-anced panel of cities


Atlantic port dummy is highly significant andthe R2 of this more restrictive regression isalmost the same as the regression reported inthe top panel

Column 2 reports estimates from an un-weighted regression The results are similar butquantitatively smaller since large Atlanticports such as London and Amsterdam nolonger get more weight Columns 3 and 4 reportweighted and unweighted estimates from simi-lar models with a dummy for potential Atlanticport that is any city that in our balanced panelcould have been used as a port for Atlantictrade The results are similar to those in col-umns 1 and 218 Column 5 drops London and

Amsterdam to show that the results are notdriven by these two major cities The coeffi-cients on Atlantic port times year interactionsare approximately halved from 1700 onwardbut they remain significant Column 6 adds afull set of country times year interactions toshow the differential growth of Atlantic portsrelative to other cities in the same country Thecoefficients on Atlantic port times year interac-tions after 1700 are about half those of column1 but still highly statistically significant

18 To allow for the specification test discussed in the textthese regressions use 1300ndash1400 as the base period Be-cause there was rapid growth in a few potentialmdashbut not

actualmdashAtlantic ports from 1400 to 1500 some of thecoefficients on potential Atlantic port are higher than thecorresponding coefficients on Atlantic port However cu-mulative growth between 1500 and any subsequent date isalways higher for Atlantic ports than for potential Atlanticports It should also be noted that some potential Atlanticports flourished as a result of secondary trade from theAtlantic

FIGURE 4A AVERAGE OF LOG CITY POPULATION IN ATLANTIC PORTS WEST EUROPEAN CITIES THAT ARE NOT ATLANTIC

PORTS AND EASTERN EUROPE (BALANCED PANEL) 1300ndash1850

FIGURE 4B AVERAGE OF LOG CITY POPULATION IN ATLANTIC PORTS MEDITERRANEAN PORTS AND WEST EUROPEAN CITIES

THAT ARE NOT PORTS (BALANCED PANEL) 1300ndash1850


Column 7 adds Asian cities from Tertius Chan-dler (1987) so now West European cities arebeing compared to both East European andAsian cities The results are similar but alsoshow the differential growth of all West Euro-pean cities relative to Asian cities19

Is there something special about ports or is itAtlantic ports that are behaving differently after1500 To answer this question Figure 4B and

column 8 show that Mediterranean ports grew atsimilar rates to inland European cities what wefind is not a general port effect but an Atlantic porteffect

Was the urban and economic expansion ofAtlantic nations driven solely by the growth ofAtlantic ports Figure 5A shows the expansionof Iberian (Spanish and Portuguese) Atlanticports other Iberian cities and West Europeaninland cities Almost all of the differentialgrowth of Spain and Portugal comes from At-lantic ports In fact non-Atlantic parts of Spainand Portugal grew more slowly than West Eu-ropean inland cities Relevant to our hypothesis

19 We also investigated the importance of the same controlsused in Table 4 for country-level growth The results whichare reported in Acemoglu et al (2002b) show that the patternin Table 5 is robust to the inclusion of these controls

TABLE 5mdashGROWTH OF ATLANTIC PORTS

Dependent variable is log city population

Balancedpanel

1300ndash1850weighted

Balancedpanel

1300ndash1850unweighted

Balancedpanel


Balancedpanel


Balancedpanel


without Londonand Amsterdam

Balanced panel1300ndash1850weighted

with full setof country

yearinteractions

Balancedpanel

weighted1300ndash1850with Asia

Balanced panelweighted

1300ndash1850with

Mediterraneanand Atlantic

ports

(1) (2) (3) (4) (5) (6) (7) (8)


Atlantic port Potential Atlantic port Atlantic port


[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)

p-value for Mediterraneanport year dummies1500ndash1850

[019]

R-squared 092 079 092 080 089 095 094 092Number of observations 1544 1544 1544 1544 1528 1544 1624 1544



[023] [004] [023] [010] [031] [033] [030] [020]

Volume of Atlantic trade Atlantic port

017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)

p-value for MediterraneanPort year dummies1500ndash1850

[014]


Notes Dependent variable is log city population from Bairoch et al (1988) Weighted regressions use current level of city population in each year as weights Allcolumns report balanced panel regressions for 1300 1400 1500 1600 1700 1750 1800 and 1850 using only cities for which we have data in all eight time periodsThe Atlantic port dummy equals one for a city used as an Atlantic port Potential Atlantic ports are all ports that could have been used for Atlantic trade and includeAtlantic ports plus ports in Belgium Germany and Ireland (there are no potential Atlantic ports in Denmark or Norway in our balanced panel) Volume of Atlantictrade is log average voyages per year this is multiplied by the Atlantic port dummy (or by the potential Atlantic port dummy) the coefficient on this interaction termis multiplied by 100 Year dummies are included for all years from 1400 Western Europe year dummies are included for all years from 1600 For a list of Atlanticports and potential Atlantic ports see the Appendix of Acemoglu et al (2002b)


below this Iberian pattern contrasts with thesteady growth of non-Atlantic British citiesshown in Figure 5B (Notice that the non-Atlantic British line starts below the West Eu-ropean line and overtakes it by 1850 seeAcemoglu et al 2002b for further evidence)

E Interpretation

The evidence presented so far has establisheda significant relationship between the potential

for Atlantic trade and post-1500 economic de-velopment and suggests that the opportunitiesto trade via the Atlantic and the associatedprofits from colonialism and slavery played animportant role in the rise of Europe This evi-dence weighs against theories linking the rise ofWestern Europe to the continuation of pre-1500trends driven by certain distinctive characteris-tics of European nations or cultures such asRoman heritage or religion

At face value this evidence is more consistent

FIGURE 5A AVERAGE OF LOG CITY POPULATION IN IBERIAN ATLANTIC PORTS OTHER IBERIAN CITIES THAT ARE NOT

ATLANTIC PORTS AND INLAND WEST EUROPEAN CITIES (BALANCED PANEL) 1300ndash1850

FIGURE 5B AVERAGE OF LOG CITY POPULATION IN BRITISH ATLANTIC PORTS OTHER BRITISH CITIES THAT ARE NOT



with theories emphasizing the direct contri-bution of profits from Atlantic trade colonial-ism and slavery such as those advancedby Williams (1944) Frank (1978) and Wal-lerstein (1974ndash1980) It is undoubtedly truethat colonial relations with the New World andAsia contributed to European growth Neverthe-less quantitative analyses for example Enger-man (1972) Engerman and OrsquoBrien (1991)OrsquoBrien (1982) and Bairoch (1993 ch 5) sug-gest that the volume of trade and the profitsgenerated by Atlantic trade appear to be toosmall to account for much of European growthdirectly Atlantic trade may also have played animportant direct role by inducing a reallocationof resources within Europe even if profits fromtrading were low (as would be the case in acompetitive economy) This direct channel isunlikely to be the whole story however sincethe volume of trade was small For exampleBairoch (1993) calculates that commodity tradebetween Western Europe and the rest of theworld amounted to less than 4 percent of theGNP of Western Europe before 1800 Althoughrecent work by Inikori (2002) argues that profitsfrom colonial activities in particular from theslave trade were larger than those estimated byOrsquoBrien even with his estimates the directeffect of Atlantic trade and colonialism couldaccount for the rise of Europe only with signif-icant increasing returns to scale in leadingsectors20

Overall therefore the weight of evidenceinclines us toward a view in which the rise ofEurope reflects not only the direct effects ofAtlantic trade and colonialism but also a majorsocial transformation induced by theseopportunities

II Our Hypothesis

A The Argument

Our hypothesis is that Atlantic trademdashtheopening of the sea routes to the New WorldAfrica and Asia and the building of colonialempiresmdashcontributed to the process of WestEuropean growth between 1500 and 1850 notonly through direct economic effects but alsoindirectly by inducing fundamental institutionalchange Atlantic trade in Britain and the Neth-erlands (or more appropriately in England andthe Duchy of Burgundy) altered the balance ofpolitical power by enriching and strengtheningcommercial interests outside the royal circleincluding various overseas merchants slavetraders and various colonial planters Throughthis channel it contributed to the emergenceof political institutions protecting merchantsagainst royal power21 Our hypothesis also im-plies that the tendency for institutional changeto emerge should have been much stronger insocieties with existing checks on royal powerthan in countries with absolutist regimes andmonarchy-controlled trade monopolies becausein these latter countries Atlantic trade did notenrich and strengthen merchant groups outsidethe royal circle as much and did not disturb thepolitical status quo

This hypothesis can be broken into 4 subhy-potheses

(a) Political institutions placing limits and con-straints on state power are essential for theincentives to undertake investments and forsustained economic growth

(b) In early modern Europe such political in-stitutions were favored by commercial in-terests outside the royal circle but were notwelcome by the monarchy and its allies

(c) Institutions favored by economically and20 For example OrsquoBrien (1982) calculates that total prof-its from British trade with less developed regions of theworld during the late eighteenth century were approxi-mately pound56 million while total gross investment during thesame period stood at pound103 million Inikori (2002 Table42) suggests that imports from the periphery around 1800were about double OrsquoBrienrsquos estimate During this periodthe aggregate savings rate was between 12 and 14 percentso if we assume that this savings rate also applies to profitsfrom trade the contribution of these profits to aggregatecapital accumulation would be less than 15 percent evenusing Inikorirsquos estimates Even assuming considerablyhigher savings rates the contribution would remain rela-tively small

21 An additional channel via which Atlantic trade mayhave contributed to institutional change may be the desire ofthe monarchy to secure the property rights of merchants inorder to encourage long-term investments in long-distancetrade Our reading of the relevant history discussed belowmakes us believe that the greater contribution of Atlantictrade to the development of capitalist institutions was bystrengthening commercial interests in favor of politicalchange in their fight against the monarchy


politically powerful groups are more likelyto prevail and

(d) In countries with nonabsolutist initial polit-ical institutions Atlantic trade and colonialactivity enriched and strengthened commer-cial interests including new groups withoutties to the monarchy

Together these four subhypotheses yieldour main hypothesis In countries with easyaccess to the Atlantic and without a strongabsolutist monarchy Atlantic trade providedsubstantial profits and political power forcommercial interests outside the royal circleThis group could then demand and obtainsignificant institutional reforms protectingtheir property rights With their newly gainedpower and property rights they took advan-tage of the growth opportunities offered byAtlantic trade invested more traded moreand fueled the First Great Divergence22

Initial institutions placing sufficient checkson the monarchy are essential for the fourthsubhypothesis so that merchants not directlyassociated with the crown benefit signifi-cantly from Atlantic trade When the power ofthe crown was relatively unchecked as inSpain Portugal and France trade was largelymonopolized and regulated the crown and itsallies became the main beneficiaries of theAtlantic expansion and institutional changedid not take place Therefore our hypothesisexplains not only the major role played byAtlantic trade in West European growth butalso why economic growth took off in Britainand the Netherlands and not in Spain andPortugal

Acemoglu et al (2002b) provide historicalevidence consistent with these subhypothesesSpace constraints preclude us from going intodetails here We refer the reader to that paperfor a more detailed discussion and briefly dis-cuss the evidence related to the fourth subhy-pothesis which is perhaps the most importantfor our argument

B Atlantic Trade and Commercial Interests

We now discuss the major changes in thepolitical institutions of Britain and the Nether-lands Our argument highlights that in bothcases (a) the political institutions at the begin-ning of the sixteenth century though not asabsolutist as in Spain and Portugal did notprovide secure property rights to commercialinterests outside the royal circle23 (b) there wassignificant conflict between these merchantgroups and the monarchy (c) Atlantic tradecreated large profits for some of these mer-chants who were in favor of institutionalchange and who then used part of these profitsto support the conflict against the crown

BritainmdashIn the British case the two mile-stones in the emergence of political institutionsconstraining royal power are the (English) CivilWar of 1642ndash1649 when Parliamentarianforces defeated Charles I and the Glorious Rev-olution of 1688ndash1689 where James II was de-posed by Parliament with the help of aninvading Dutch army and replaced by Williamof Orange and a parliamentary regime with aconstitutional monarchy Although there is noconsensus among historians on the relative im-portance of these two events this is secondaryfor our focus What is important is that therewas a major improvement in British politicalinstitutions between the mid-seventeenth andearly-eighteenth centuries

Although after the War of the Roses Britainwas never as absolutist as France Portugal andSpain both the Tudor and Stuart monarchs con-sistently attempted to expand their powers Theinsecurity of property rights was clear duringthe reign of Henry VIII when there were con-tinual attempts to regulate trade and underminethe powers of Parliament (see Geoffrey R El-ton 1991) A significant attempt to establish aform of absolutism came during the period ofso-called ldquopersonal rulerdquo of Charles I after hedissolved his third Parliament in 1629 raisedtaxes in an unconstitutional way and used the

22 The establishment of political institutions limiting thepower of the monarchy may have also created positivespillovers in the rest of the economy especially for indus-trial capitalists (consistent with the subsequent growth ofnon-Atlantic British cities in Figure 5B)

23 More explicitly these commercial interests includedmerchants not receiving crown-granted monopolies slavetraders various producers in the colonies and parts of thegentry in Britain and the majority of the Dutch merchantsnot allied with the Habsburg monarchy in the Netherlands


Star Chamber to manipulate legal decisions inhis favor (Kevin Sharpe 1992)

Although undoubtedly complex socialevents both the Civil War and the GloriousRevolution were also battles over the rights andprerogatives of the monarchy24 In both casescommercial interests (including large segmentsbut not all of the merchants and the gentry)predominantly sided with those demanding re-strictions on the power of the monarchy Duringthe Civil War for example the majority of themerchants and even many of those with royalmonopolies supported Parliament (see Brenner1973 2003 Mary Keeler 1954 Douglas Brun-ton and D H Pennington 1954)25 Members ofthe Commons from the City of London whichwas the main center of mercantile activity aswell as many non-London commercial constit-uencies such as Southampton Newcastle andLiverpool supported Parliament against theKing David H Sacks (1991 pp 230ndash47)shows that in Bristol trading commercial andindustrial interests outside of the Merchant Ad-venturers (the trading company then enjoyingthe royal monopoly) were ParliamentariansBrunton and Pennington (1954 p 62) also notethat ldquoin the country as a whole there was prob-ably a preponderance of Parliamentarian feelingamong merchantsrdquo

The situation for the Glorious Revolution is

similar The East India Company under the con-trol of Josiah Child supported James II hisclaim to tax without consent of Parliament andhis right to grant trading monopoliesmdashof whichit was the main beneficiary But the majority ofcommercial interests alienated by James IIrsquosgrants of various monopoly privileges and es-pecially the interlopersmdashmerchants trying tobreak into trade with Asiamdashwere on the side ofthe revolution (Bruce G Carruthers 1996 Pin-cus 2002) These merchants also receivedstrong support from Whigs who sought to con-strain the king (Henry Horwitz 1978) Summa-rizing the evidence Pincus (2002 p 34)concludes ldquoEnglandrsquos merchant community ac-tively supported Williamrsquos plan for invasionand provided a key financial prop to the regimein the critical early monthsrdquo

The victory of Parliament in the Civil Warand after the Glorious Revolution introducedmajor checks on royal power and strengthenedthe rights of merchants After the Civil War thefraction of MPs who were merchants increaseddramatically Although even in the 1690s thisnumber was not large enough to constitute amajority on its own as David Stasavage (2003)shows the interests of merchants were assuredby the formation of the Whig coalition of mer-chants and Protestant landowners This periodalso witnessed a series of policies favoring mer-chants including the Navigation Acts of 1651and 1660 which restricted trade with Britishcolonies to British ships and merchants (J EFarnell 1964 J P Cooper 1972) and strength-ened the position of British overseas tradersespecially slave traders (see Geoffrey Holmes1993 p 64) Similarly the Glorious Revolutionled to a series of economic reforms sought bymerchants outside the royal circle including thedismantling of all monopoly charters except theEast India Company (Perry Gauci 2001) andthe establishment of the Bank of England Theconventional wisdom in economic history em-phasizes the importance of these institutionalchanges for the protection of property rightsand how they led to a wave of innovations ineconomic institutions particularly in financialmarkets (eg North and Weingast 1989 Car-ruthers 1996 Larry Neal 2000)

Critically for our thesis the major changes inpolitical institutions and the new assertivenessof merchant groups coincided with the expan-sion of British mercantile groups trading

24 Other prominent interpretations of the English CivilWar have emphasized various factors apart from those westress here Conrad Russell (1990) argues that the Civil Warwas a plot by the traditional aristocracy to regain power ithad lost under the Tudors Many for example John SMorrill (1993) focus on the role of religious differences indetermining who supported which side and recent work byBrian Manning (1996) stresses more general class conflictAlthough there are doubtless elements of truth in theseapproaches the general role of mercantile interests seemsundeniable (see Roger C Richardson 1998 for a balancedoverview of the debate)

25 Valerie Pearlrsquos seminal study (1961) argued that therewere political divisions between such groups as the Mer-chant Adventurers who benefited from monopolies grantedby the crown and new merchants who did not For exam-ple the two prendashCivil War MPs for Bristol HumphreyHooke and Richard Long were Royalists Robert Ashton(1979 1996) on the other hand documented that evenmerchants who enjoyed monopolies tended to oppose thecrown by the time of the Civil War and argued ldquothemajority of the City fathers far from being the naturalsupporters of Stuart absolutism at the end of the period ofCharles Irsquos personal rule in the late 1630rsquos were as alien-ated from royal policies as were the vast majority of thepolitical nationrdquo (1996 p 3)


through the Atlantic The East India Companywas founded in 1600 and from 1600 to 1630there was an unprecedented wave of investmentby merchants gentry and even some aristoc-racy in overseas ventures (Theodore K Rabb1967) Virginia tobacco cultivation boomed inthe 1620s and beginning in the 1640s thehighly profitable Caribbean sugar colonies de-veloped Finally in the 1650s the British beganto take over the Atlantic slave trade

A number of historians most notably Bren-ner (2003) have emphasized that Atlantic mer-chants were critical in ensuring the militaryvictory of Parliament in the Civil War In hisfamous book Lawrence Stone also points outthat ldquo other important merchant elements cannow be identified men interested especially inthe American trades in New England coloniza-tion and in breaking the monopoly of the EastIndia and Levant Companies They were newmen in new fields of entrepreneurial endeavorwho chafed at the political and economic stran-glehold of the older established monopolisticoligarchies These men were important mem-bers of the group of radicals who seized controlof London at a critical moment in 1641 and soswung the power and influence of the city de-cisively on the side of Parliamentrdquo (1973 p144)

Atlantic trade indeed created large profits forthe fortunate who succeeded in this high-riskendeavor In the Appendix we use data onprofits from K N Chaudhuri (1965) and deVries and van der Woude (1997) on investmentfrom Rabb (1967) on trade from Inikori (2002)and de Vries (2003) and on rates of return fromRichard Grassby (1969 1995) and OrsquoBrien(1982) to estimate profits of various merchantcompanies slave traders and colonial plantersfrom Atlantic trade during this period Theseestimates suggest that profits from Atlantictrade were negligible before 1575 aboutpound40000 on average per annum from 1576 to1600 (mostly from a few highly profitable pri-vateering expeditions) perhaps pound200000 on av-erage per annum from 1601 to 1650 and aroundpound500000 per annum from 1651 to 1675 Profitsthen rose with the expansion of sugar and theslave trade to around pound900000 per annum from1676 to 1700 pound17m per annum from 1701 to1750 and probably about pound5m per annum in thelate eighteenth century (all figures adjusted to1600 prices using the index of building crafts-

menrsquos wages from E H Phelps Brown andSheila V Hopkins 1955) These profits weresubstantial relative to the personal wealth ofmerchants and gentry during this time periodFor example personal wealth of pound10000 in theearly seventeenth century was enough to bevery rich A minimum investment of pound2000was required to become a director of the EastIndia Company and pound200 represented a sub-stantial investment (Brenner 1973 pp 62ndash63Brenner 2003 p 78) Moreover because prof-its from Atlantic trade were highly concen-trated they created a number of very wealthymerchants (see Grassby 1995 pp 248 and263) These profits were also large relative tothe resources necessary to make a differencepolitically and militarily26

Many merchants used their profits from At-lantic trade to support the conflict against thecrown For example the Earl of Warwick whoearned at least pound50000 from privateering in oneyear prior to the Civil War (W Frank Craven1930) applied his fortune and experience withnaval warfare to effectively oppose the king27

More generally Parliament during the CivilWar was partly financed by taxes on and profitsfrom Atlantic trade Parliamentary leaders suchas Sir Edwin Sandys and John Pym were activein colonization and trade with the AmericasJames I well aware of the links between majorAtlantic trading ventures and parliamentary

26 From 1550 to 1688 the English monarch was alwaysshort of cash and war typically required additional fundsFor example King Charles I was forced to recall Parliamentin 1640 because he needed to raise about pound300000(pound250000 in 1600 prices) for war against Scotlandmdashenough to pay and equip about 12000 soldiers for a year(Ashton 1960 p 176) Total English government revenueswere around pound500000 in 1600 and about pound850000 in 1640(Richard Bean 1973) Armies on both sides of the EnglishCivil War were small 10000 to 20000 men and most ofthe conflict was small-scale local operations by regionalforces (Geoffrey Parker 1988 pp 28 and 41) Parliamentfielded 27000 at the battle of Marston Moor and just 13000at Naseby the presence or absence of a few thousand troopswas therefore decisive (Parker 1988 p 41) Kennedy(1987 p 63) estimates that the average annual cost of asoldier was around pound27 in 1657 (about pound20 in 1600 prices)

27 A privateer is an armed private vessel bearing theauthorization or commission of a sovereign power to attackan enemy ie a privately funded and manned extension ofa countryrsquos naval forces Privateers typically engaged intrading activities as well as fighting (see for example thecase of John Hawkins discussed in N A M Rodger 1997p 201)


opposition intervened in the election of trea-surer for the Virginia Company sayingldquoChoose the devil if you will but not Sir EdwinSandysrdquo (Rabb 1998 p 349) Similarly for theGlorious Revolution Pincus (2002 pp 32ndash33)provides evidence that ldquothe merchant commu-nity poured money into William of Orangersquoscoffers in 1688rdquomdashperhaps around pound800000(about pound500000 in 1600 prices) enough to payfor a sizable army

The NetherlandsmdashDutch merchants alwayshad considerable autonomy and access to prof-itable trade opportunities Nevertheless prior tothe Dutch Revolt the Netherlands (in fact theentire Duchy of Burgundy) was part of theHabsburg Empire and the political power ofDutch merchants was limited The Habsburgmonarchy consistently attempted to increase itspolitical dominance over and fiscal revenuesfrom the Netherlands (W Fritschy et al 2001)The critical improvement in Dutch political in-stitutions was therefore the establishment of theindependent Dutch Republic with politicaldominance and economic security for mer-chants including both the established wealthyregents and the new merchants immigratingfrom Antwerp and Germany28

Dutch politics was shaped by the conflictbetween Dutch merchants and the Habsburgmonarchy starting in the fifteenth century andbefore then by the conflict between merchantsand the Duke of Burgundy By 1493 Maximil-ian of Habsburg had reversed the Grand Privi-lege of 1477 which gave the states general theright to gather on their own initiative and curbedthe right of the ruler to raise taxes After 1552war with France and England increased theHabsburgsrsquo fiscal needs and led them to imposea large tax burden on the Netherlands Growingfiscal and religious resentment in 1572 led to aseries of uprisings mostly orchestrated by com-mercial interests (see Jonathan I Israel 1995)These culminated in a war of independencewhich began with the Revolt in the 1570s anddid not end until 1648 punctuated by Philip IIdiverting resources to intervene in France after

1590 the successful Dutch offensives of 1591ndash1597 under the command of Maurice of Nassauthe embargoes against Dutch trade with Spainand Portugal in 1585ndash1590 1598ndash1609 and1621ndash1647 and the Twelve Years Truce from1609 to 1621

The major turning point came in the 1590swhen important changes in Dutch military andcommercial strategy became evident New mil-itary tactics made it possible for the Dutch tohold their own against experienced Spanish in-fantry (Geoffrey Parker 1988 pp 19ndash20) Thiswas combined with a fiscal and financial ldquorev-olutionrdquo that allowed states particularly Hol-land both to increase their tax revenues andborrow against future taxes in order to financethe war effort (Fritschy 2003) At the sametime the Dutch took the critical strategic step ofseeking direct access to Asian and Americantrade centers This both enriched a generation ofDutch merchants and undermined Spanish andPortuguese revenues sufficiently to inducePhilip III to offer peace By 1605 it was clear toa Spanish royal councillor the Count of Oli-vares that victory would go to ldquowhoever is leftwith the last escudordquo (Parker 1977 p 238)

Merchants were naturally the primary politi-cal and economic force on the side of indepen-dence De Vries and van der Woude (1997)argue that ldquourban economic interests ultimatelybelieved it advantageous to escape theHabsburg imperial frameworkrdquo (p 369) Theyalso note that in the case of Amsterdam theldquo[Habsburgsrsquo] opponents included most of thecityrsquos international merchants [I]n 1578 anew Amsterdam city council threw the cityrsquos lotin with the Prince of Orange among the mer-chants returning from exile were [thosewhose families] and several generations of theirdescendents would long dominate the cityrdquo(1997 p 365)

Commercial interests involved in the Atlanticwere particularly important in the shaping of theconflict (see for example Israel 1982 1995Herman van der Wee 1993 pp 272ndash73) In1609 in an attempt to prevent the creation ofthe Dutch West India Company Philip III of-fered peace and independence in return for aDutch withdrawal from both the West and EastIndies But these terms were ldquosimply not feasi-ble politically because many regents and elitemerchants had invested heavily in the [DutchEast India Company]rdquo (Israel 1995 p 402)

28 By the year 1600 a third of the population of Amster-dam was immigrants (Israel 1995 p 309) In 1631 therewere 685 citizens of Amsterdam with wealth over 25000florins Only half of them were native Hollanders (Parker1977 p 251)


Prominent in the anti-peace camp was the fa-mous Dutch leader and general Maurice of Nas-sau who was heavily involved in colonialtrades and ldquoReynier Pauw the preeminent fig-ure and leader of the anti-truce faction in Hol-land who besides being a champion of the WestIndia Company project had been a foundermember of the East India Company and formany years a director of its Amsterdam cham-berrdquo (Israel 1982 p 40)

It is therefore no surprise that independenceput merchants firmly in control of the politicalprocess De Vries and van der Woude (1997 p587) describe the new political elite followingthe Dutch Revolt as ldquo6 to 8 of urban house-holds with incomes in excess of 1000 guildersper year This was the grote burgerij fromwhom was drawn the political and commercialleadership of the country Here we find first andforemost the merchantsrdquo They also point outhow merchants dominated the governments ofLeiden Rotterdam and the cities in the twolargest states Zeeland and Holland

The Dutch economy had been expandingsince the fifteenth century and experienced ad-vances in economic institutions including inshipping agriculture and finance particularlypublic finance prior to this revolt (James DTracy 1985 Jan Luiten van Zanden 1993)Nevertheless the potential of these institutionswas severely limited under the Habsburg yokebecause of the threat of arbitrary taxation Forexample Marjolien Hart et al (1997 fig 23 p19) show that despite the changes in financialinstitutions in the mid-sixteenth century inter-est rates did not fall systematically until after1600 when they declined to about one-third oftheir pre-revolt level Consequently the econ-omy appears to have experienced a major trans-formation after the process of political changebegan Van Zanden (1993) notes ldquoWe can seethe starting point of the rapid urbanization at1580rdquo (pp 35ndash36) and continues ldquoduring thistransformation process the pre-1580 proto-capitalist structure disappeared out of thislsquounspecialisedrsquo class of small-holders fisher-men homeworkers and sailors separate classesof large farmers agricultural laborers andcraftsmen aroserdquo (p 39) Similarly Braudel(1995 p 547) dates the start of the divergencebetween the South and North of Europe to 1590with the ldquoexplosionrdquo of Dutch commerce andthe rise of Amsterdam

Critical was the Dutch merchantsrsquo improv-ing economic fortunes partly from Atlantictrade which were used to field a powerfularmy against the Habsburg Empire The Bal-tic trade is widely recognized as important forthe Dutch economy in the sixteenth centurybut profits from Atlantic trade quickly sur-passed those from Baltic trade and providedthe funds necessary for the Dutch militaryeffort against the Habsburgs (Israel 1989)De Vries and van der Woude (1997) estimatethat the annual profits of the Dutch East IndiaCompany alone between 1630 and 1670 21million guilders per annum were more thantwice the total annual profits from the Balticgrain trade between 1590 and 1599 (pp 373and 447)

Fritschy (2003) estimates that as a result ofthese developments tax revenue per head inHolland rose nearly fivefold from 1575 to1610 while population increased by a third(see also Tracy 2001 Table 72) These rev-enues enabled Holland to provide 960000guilders for the war in 1579 and to pay fivemillion guilders in 1599 (Parker 1977 p251) Israel (1995 pp 241ndash 42) summarizesthe basic reason for the Dutch victory asfollows ldquoFrom 1590 there was a dramaticimprovement in the Republicrsquos economic cir-cumstances Commerce and shipping ex-panded enormously as did the towns As aresult the financial power of the states rapidlygrew and it was possible to improve the armyvastly both qualitatively and quantitativelywithin a short space of time The army in-creased from 20000 men in 1588 to 32000by 1595 and its artillery methods of trans-portation and training were transformedrdquo By1629 the Dutch were able to field an army of77000 men 50 percent larger than the Span-ish army of Flanders (Israel 1995 p 507)

Overall both the British and Dutch evidencetherefore appears favorable to our hypothesisthat Atlantic trade enriched a group of mer-chants who then played a critical role in theemergence of new political institutions con-straining the power of the crown

Spain Portugal and FrancemdashThere is gen-eral agreement that Spanish and Portuguese po-litical institutions at the turn of the sixteenthcentury were more absolutist than those in


Britain and the Netherlands and did not expe-rience similar reform29

A key difference between these cases and theBritish-Dutch patterns is the organization oftrade which in turn reflected differences inpolitical institutions Throughout this periodthe granting of trade monopolies was a centraltool for the rulers to raise revenue When thepower of the monarchs was constrained theywere unable to use this fiscal tool For examplethe English Parliament successfully blockedmany attempts of both Tudor and Stuart mon-archs to create such monopolies (ChristopherHill 1969) Consequently in Britain ldquomosttrade was carried on by individuals and smallpartnerships and not by the Company of Mer-chant Adventurers the Levant Company orothers of their kindrdquo (Davis 1973a p 41) Atleast by 1600 there was quite free entry into theBritish merchant class (R G Lang 1974) Incontrast Rondo Cameron (1993 p 127) de-scribes the Portuguese situation as followsldquoThe spice trade in the East Indies of the Por-tuguese Empire was a crown monopoly thePortuguese navy doubled as a merchant fleetand all spices had to be sold through the Casada India (India House) in Lisbon no com-merce existed between Portugal and the Eastexcept that organized and controlled by thestaterdquo (See also Charles R Boxer 1985 Earl JHamilton 1948) Similarly in Spain colonialtrade was a monopoly of the Crown of Castilleand was delegated to the Casa de Contratacion(House of Trade) in Seville which was itselfclosely monitored by the government (James HParry 1966 ch 2)

France on the other hand can be viewed asan intermediate case Although French institu-tions were equally absolutist (W F Church1969 David Parrott 2001) early Atlantic ac-tivity enriched some merchant groups in par-ticular the protestant Huguenots However themonarchy soon clashed with and defeated theHuguenots first with the siege of La Rochelleby Louis XIII and then the outlawing of theProtestant church by Louis XIV (see eg War-ren C Scoville 1960) The monarchy then keptmuch of overseas trading activity as a royalmonopoly especially under Colbert (see egDavis 1973b pp 222ndash24 William Doyle1974 pp 210ndash11) Nevertheless certain strongFrench commercial and industrial interests de-veloped and arguably forced institutionalchange before during and after the FrenchRevolution (see G Lefebvre 1947 and Doyle1988 for the debate on the origins of the FrenchRevolution)

Overall the evidence is therefore consistentwith our thesis that in Spain and Portugal andalso largely in France merchant interests withsufficient power to challenge the crown did notdevelop because the crown and groups allied toit were the main beneficiaries of the profitsfrom transoceanic trade and plunder

III Atlantic Trade and Institutional Change

We now attempt to substantiate our hypoth-esis further by providing empirical evidence onthe link between changes in political institutionsand Atlantic trade A prerequisite for this exer-cise is a measure of relevant political institu-tions Unfortunately no such measure exists forthis period30 So as a first step we attempted tocreate a measure of political institutions forEuropean countries between 1300 and 1850adapting the definition of ldquoconstraint on theexecutiverdquo from Gurrrsquos Polity dataset This is auseful concept since it measures limitations onthe arbitrary use of power by the executive (forthe relevant time period the monarchy) and is

29 Davis (1973a p 66) for example emphasizes thehigh degree of absolute control by the monarchy in Spain asfollows [in Castille] ldquothe king ruled subject only to weakconstitutional restraints In the first decades of the sixteenthcentury the crown had reduced the pretensions of the Castil-lian nobility and towns so that the representative body theCortes could obstruct but not in the last resort prevent royaltax raisingrdquo and contrasts this with the situation in Britain(eg Davis 1973a p 210)

The modern literature in particular I A A Thompson(1994) and Michael A R Graves (2001) suggests that theextent of Spanish absolutism has been overemphasized byscholars such as North and Thomas (1973) and points outimportant differences between Castille and such other partsof Iberia as Aragon and Catalonia Nevertheless it is cer-tainly true that the Spanish Crown was able to create trademonopolies and raise taxes in ways that the Tudor andStuart monarchies could not

30 An alternative approach would be to use the terms offinance for trading entities as suggested by one of ourreviewers Although there are some data on this the cov-erage is not broad enough for our purposes Moreover theterms of finance may be affected by the demand for andsupply of financial resources as well as the underlyingsecurity of property rights


presumably correlated with the security of prop-erty rights for merchants and the control overthe monopoly of overseas trade by themonarchy31

We follow the Polity IV coding handbookgiving a score of between 1 and 7 for constrainton the executive to each country32 For 1800and 1850 we use the Polity coding for con-straint on the executive where available Forearlier periods we coded these measures our-selves The main source for this exercise wasWilliam L Langer (1972) a classic historicalencyclopedia written with a focus on constitu-tional events We supplemented this work withthe more recent edition by Peter N Stearns(2001) While there may be disagreement aboutthe precise values used in particular years thegeneral level of constraint on the executive doesnot appear to be controversial For example theabsolutist regimes of France Portugal andSpain clearly had much less constraint on theexecutive than did the Netherlands after inde-pendence or England after the Civil WarAcemoglu et al (2002b) give further details andreport the entire series

Table 6 documents the differential changesin institutions between Atlantic traders andother West European nations by estimating anequation similar to (1) with constraint on theexecutive as the left-hand-side variable The

results show significant differential improve-ments in institutions among Atlantic tradersand no evidence of differential existingtrends Unlike our results when urbanizationwas the dependent variable however evenafter the inclusion of Atlantic trade interac-tions there is some evidence of differentialWest European effects

Other columns use the same controls andtime interactions as in Table 4 Although theF-statistics show that many of these time inter-actions are significant neither Protestantismnor wars nor Roman heritage nor latitude ap-pears to have led to greater institutional changeafter 1500 (for example institutions in Protes-tant countries improved more rapidly until1750 and significantly more slowly thereafter)

Overall these results suggest that followingthe surge in Atlantic trade there were greaterstrides toward better political institutions in na-tions engaged in Atlantic trade and colonialism(or in those with a greater potential to engage inAtlantic trade)

IV The Role of Initial Institutions

We now investigate whether as implied byour hypothesis it was predominantly societ-ies with less absolutist initial institutions (andrelatedly those without widespread royalgranted monopoly rights in overseas trade)that took advantage of the opportunities of-fered by Atlantic trade We also investigatethe related hypothesis of North and Thomas(1973) and Jones (1981) that post-1500 de-velopments largely reflect divergence be-tween societies that had very differentpolitical institutions at the turn of the fifteenthcentury This differs from our hypothesiswhich emphasizes the interaction betweeninitial political institutions and Atlantic trade

To investigate these ideas we estimate mod-els of the following form

(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt

ln ATt PATj Ij1415 jt

31 The measure of constraint on the executive may not beideal for our purposes however since a number of signif-icant constraints on monarchs were imposed by the noblesand did not necessarily serve to protect the rights of mer-chants For example in much of the 1500ndash1750 periodPoland had a highly constrained executive But there wasrelatively little protection for urban merchants most of therights rested with the nobility For this reason we modifiedthe definition of constraint on the executive to create analternative measure which we refer to as ldquoprotection forcapitalrdquo The coding of this measure depends on the formalrights given to urban merchants particularly their protectionin the event of a dispute with the nobility or monarch Theresults using this measure are similar to those using con-straint on the executive and are contained in Acemoglu etal (2002b)

32 A value of 1 means ldquothere are no regular limitations onthe executiversquos actionsrdquo 3 means ldquothere are some real butlimited restraints on the executiverdquo 5 means ldquothe executivehas more effective authority than any accountability groupbut is subject to substantial constraints by themrdquo and 7means ldquoaccountability groups have effective authority equalto or greater than the executive in most activityrdquo Scores of2 4 and 6 are used for intermediate values See Monty GMarshall and Keith Jaggers (2000)


where as before ujt is the urbanization rate ln ATtis our measure of Atlantic trade PATj is againeither a dummy for Atlantic trader or the Atlanticcoastline-to-area ratio and Ij1415 is country jrsquosinitial institutions calculated as the average of itsconstraint on the executive in 1400 and 1500 Wechoose the average of these two dates to capturethe long-term institutional differences in the pre-1500 period The t Ij1415 dt terms allow anydifferential economic trends related to differencesin initial institutions that would apply even withno access to the Atlantic Significant coefficientson these interaction terms would imply that atleast part of the post-1500 developments in Eu-rope reflect divergent paths taken by countrieswith different initial institutions independent ofthe effects of Atlantic trade The table reports thep-value from a joint significance test for all of

these interaction terms The ln ATt PATj term onthe other hand measures the effect of Atlantictrade for a given level of institutions In the tablethis term is evaluated at the lowest score of insti-tutions ie for Ij1415 1 so the coefficient onthis term measures the growth contribution ofAtlantic trade and access to the Atlantic for asociety with the worst possible initial institutions

The variable ln ATt PATj Ij1415 tests thehypothesis of interest A significant coefficient implies that there were divergent paths takenby countries with different initial institutionsbut this divergence relates significantly towhether they took advantage of the opportuni-ties presented by Atlantic trade

The results are reported in Table 7 using theAtlantic trader dummy for the potential for At-lantic trade PATj (results using the coastline-

TABLE 6mdashATLANTIC TRADE AND INSTITUTIONS

Dependent variable is constraint on the executive

Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850


Panel1300 to 1850using Atlantic

coastline-to-areameasure of

potential forAtlantic trade




(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)

Potential for Atlantic trade 042 045 043 039 043 1299 volume of Atlantictrade

(006) (006) (006) (006) (006) (231)

p-value for Protestant year effect

[000]

Wars per year in precedingcentury

0034(020)


[005]

p-value for latitude year [049]R-squared 075 085 081 084 081 082 081 081 079Number of observations 192 192 192 192 176 192 192 192 192

Notes Standard errors are in parentheses Weighted panel regressions with full set of country and year dummies Weights are total population in each country in eachyear from McEvedy and Jones (1978) Dependent variable is constraint on executive which ranges from 1 to 7 where a higher score indicates more constraints onarbitrary action by the executive All columns use the Atlantic trader dummy (one for Britain France Spain Portugal and the Netherlands zero for all others) as themeasure of potential for Atlantic trade apart from columns 8 and 9 which use the ratio of Atlantic coastline to area (including Atlantic traders plus Belgium DenmarkGermany Ireland and Norway) Volume of Atlantic trade is the log average number of voyages per year Protestant is a dummy for whether country was majorityProtestant in 1600 Protestant year is the Protestant dummy interacted with year dummies for 1600 and after Wars per year are for the preceding century through1700 1700ndash1750 for 1750 1750ndash1800 for 1800 and 1800ndash1850 for 1850 Roman heritage is dummy for whether country was in the Roman Empire this is interactedwith year dummies for 1600 and after Latitude is distance from the equator for capital city of this country today this is interacted with year for 1600 and after


to-area ratio measure are identical and arecontained in Acemoglu et al 2002b) Panel Apresents estimates of equation (3) while panelB presents estimates of a similar equation withlog income per capita as the dependent variablePanel C shows the role of the interaction be-tween initial institutions and Atlantic trade forthe evolution of institutions

The results in all three panels are similar Theinteraction between the aggregate measure ofAtlantic trade and potential for Atlantic trade ln

ATt PATj is generally significant by itself andalso when entered against the t Ij1415 dtterms This shows that the ability to take advan-tage of Atlantic trade was of major impor-tance for post-1500 developments When weadd the triple interaction ln ATt PATj Ij1415 this is typically the only significant term33

33 When the ln ATt PATj Ij1415 term is included lnATt PATj has typically a negative and sometimes significant

TABLE 7mdashINTERACTION BETWEEN INITIAL INSTITUTIONS AND ATLANTIC TRADE

Using Atlantic trader dummy as measure of Atlantic trade

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

Panel A Dependent variable is urbanizationPanel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850


Atlantic trader dummy volume of Atlantic trade

0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)

p-value for initial institutions year (1600 1700 17501800 1850)

[061] [051] [071] [085] [012] [008] [042] [092]

Volume of Atlantic trade 0021 0017 0021 0022initial institutions Atlantic trader dummy

(0004) (0005) (0004) (0004)


Panel B Dependent variable is Log GDP per capitaPanel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)


Panel C Dependent variable is constraint on the executivePanel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)


Notes Standard errors are in parentheses Weighted panel regressions with full set of country and year dummies Weights are total population in each country in eachyear from McEvedy and Jones (1978) Dependent variable is urbanization in panel A log GDP per capita in panel B and constraint on the executive in panel CWestern Europe dummies interacted with years (from 1600) are included in all columns but not reported to save space Urbanization in Europe is from Bairoch etal (1988) and urbanization in Asia is from Bairoch (1998) Log GDP per capita is from Maddison (2001) Constraint on the executive is coded from Langer (1972)initial institutions are the average of institutions in 1400 and 1500 We use the Atlantic trader dummy as the measure of potential for Atlantic trade Volume of AtlanticTrade is the log average number of voyages per year and is demeaned Main effects are evaluated at initial institutions equal to one For data definitions and sourcessee Appendix Table 1


For example the coefficient of 0021 on this tripleinteraction term in column 4 implies that urban-ization in an Atlantic trader with an initial con-straint on the executive equal to 3 like theNetherlands grew by 157 percentage points morethan urbanization in an Atlantic trader countrywith the worst initial institutions 1 (0021 2 374 0157 where 374 is the change in the logvolume of Atlantic trade between 1500 and 1800)

These results imply that the patterns reportedso far are explained almost entirely by the factthat countries with initially constrained rulerswere able to take advantage of the opportunitiespresented by Atlantic trade Although Spain andPortugal benefited from the transfer of resourcesfrom the New World during the sixteenth centurythey neither developed the political institutionsto support economic growth nor experiencedsustained economic development Our evidencesuggests that these differential patterns areclosely related to the fact that they started thepost-1500 era with absolutist regimes in controlof overseas activity On the other hand it ap-pears that the Italian city-states which startedwith relatively nonabsolutist institutions around1500 did not experience further economic de-velopment because they did not have as easyaccess to the Atlantic as Britain and the Neth-erlands did Britain and the Netherlands werethe economic winners because they had bothrelatively good political institutions to start withand ready access to the Atlantic

V Conclusion

This paper documents a distinctive and inter-esting fact related to the process of Europeangrowth between 1500 and 1850 the growth ofnations with access to the Atlantic and thegrowth of Atlantic ports account for most of thedifferential growth of Western Europe relativeto Eastern Europe It therefore appears that therise of Europe between 1500 and 1850 waslargely the rise of Atlantic Europe and the riseof Atlantic ports This fact weighs against the-ories of the origins of European development

emphasizing distinctive European characteris-tics and purely internal dynamics but is consis-tent with those that give a prominent role toAtlantic trade and deemphasize the continuationof pre-1500 trends or permanent European char-acteristics such as religion Roman heritage orEuropean culture If these factors are importantit must be because of the interaction betweenthem and the opportunity to trade in theAtlantic

We suggested that Atlantic trade contrib-uted to European growth through an indirectinstitutional channel as well as via its moreobvious direct effects Our hypothesis is thatAtlantic trade generated large profits for com-mercial interests in favor of institutionalchange in countries that met two crucial pre-conditions easy access to the Atlantic andnonabsolutist initial institutions These profitsswung the balance of political power awayfrom the monarchy and induced significantreforms in political institutions which intro-duced more secure property rights and pavedthe way for further innovations in economicinstitutions With their newly gained propertyrights English and Dutch merchant nationsinvested more traded more and spurred eco-nomic growth

Our analysis stopped before West Europeanindustrialization focusing instead on economicand political developments between the six-teenth and nineteenth centuries Consequentlywe did not investigate why some successfulAtlantic nations like the Dutch did not indus-trialize early while Britain and some non-Atlantic nations such as Germany did Wesuspect that the answer is related to interstatecompetition ldquodefensive modernizationrdquo re-sponses of certain European nations and pos-sibly the adverse effects of oligarchies onindustrialization but we leave further investiga-tion of this issue for future research

The process of early modern Europeangrowth is undoubtedly multifaceted We areaware that our account leaves out many impor-tant aspects of the social and economic devel-opment of Western Europe Our intention is notto offer a mono-causal explanation for the riseof Europe but rather to suggest that Atlantictrade played a major role in this process It isour hope that our hypothesis and the empiricalpatterns documented in this paper will encour-age further research

coefficient reinforcing the conclusion that nations withabsolutist institutions did not benefit much or at all fromthe opportunity to trade in the Atlantic In addition in threespecifications in Table 7 the interactions between initialinstitutions and dates after 1600 are jointly significant butthe coefficients (not shown in the tables) are negative


APPENDIX CONSTRUCTION OF KEY VARIABLES

Country-Level and City-Level UrbanizationDatamdashCalculated from the urban populationdataset of Bairoch et al (1988) and countrypopulation estimates from McEvedy and Jones(1978) Details are provided in the Appendix ofAcemoglu et al (2002b)

Trade MeasuresmdashAcemoglu et al (2002b)explain in detail the construction of Atlanticand Mediterranean trade volume measuresThese series are annual average voyagesequivalent for ships of 400 deadweight tonsThe Mediterranean trade estimates are basedon information on Venetian trade levels fromFrederic Chapin Lane (1934) but we alsoinclude Genoa Catalonia and other tradingcenters (Carla Rahn Phillips 1990) Estimatesexclude short-haul coastal trade and trade bythe British and Dutchmdashthese countries alsoengaged in Mediterranean trade as they builttheir naval power and trading empires after1600

Key sources for our Atlantic trade seriesare de Vries (2003) Tracy (1990) Davis(1962) and N Steensgaard (1974) We havealso constructed an alternative Atlantic tradeseries based on Kevin H OrsquoRourke and Jef-frey G Williamson (2002) Robustness re-sults using this series are reported inAcemoglu et al (2002b) The growth of ourvolume-based Atlantic trade series matchesclosely the sum of annual value of Europe-Africa-New World commerce series in Inikori(2002 Table 48 p 202) and de Vriesrsquo (2003)trade flows with Asia

Estimates of British Profits from TrademdashAllfigures are approximately in 1600 prices usingthe index of building craftsmenrsquos wages con-structed by Phelps Brown and Hopkins (1955)which shows a doubling of wages from 1500 to1600 then a 50 percent increase from 1600 to1650 followed by rough stability through 1700and a further 50 percent increase during theeighteenth century

1576ndash1600mdashRabb (1967 pp 61ndash62) calcu-lates that total profits from privateering in1585ndash1603 were pound700000 Dividing by 25years gives an average of pound28000 per yearapproximately pound40000 in 1600 prices

1601ndash1650mdashProfits for the vertically inte-grated Dutch East India Company from1630 ndash1670 were 21m guilders (de Vries andvan der Woude 1997 p 447) British tradewith Asia was around a half of Dutch levels inthe seventeenth century (de Vries 2003) andthe guilder-pound exchange rate fluctuatedaround 10 so total British profits from Asiantrade (including interlopers and suppliers)were likely around pound100000 per annum(which is consistent with Chaudhuri 1965)Around pound10m was invested between 1600 and1630 in joint stock companies active in theNew World and Africa (Rabb 1967) Evenwhen a company failed to show returns aswith the Virginia Company individual colo-nists and their suppliers could earn good prof-its Privateering in the 1630s and 1640s washighly profitable (Craven 1930) We assumethe same level of earnings in the New Worldas in East India trade ie pound100000 per an-num yielding an estimate of average annualprofits of pound200000 in 1600 prices

1651ndash1675mdashFrom 1650 we use the annualvalue of export production in British Americafrom Inikori (2002 p 181) This waspound421000 in 1651ndash1670 and pound27m per annumfor 1711ndash1760 we take the average value for1651ndash1700 to be pound1m OrsquoBrienrsquos (1982) num-bers suggest that profits were 50 percent ofimport volume implying profits of pound500000To this we add pound100000 per annum fromthe East India trade with the same calcula-tion as above yielding profits of pound600000per annum or approximately pound500000 in 1600prices

1676ndash1700mdashInikorirsquos British Americatrade estimate is pound27m per annum for 1711ndash1760 we assume pound2m per annum for 1676ndash1700 which implies profits of pound1m AddingEast India profits of pound100000 gives an averageannual profit estimate of pound11m or pound900000 in1600 prices

1701ndash1750mdashInikorirsquos British America tradeestimate is pound68m for 1761ndash1780 we take theaverage value for 1701ndash1750 to be pound4m thusprofits of pound2m Adding again profits ofpound100000 from East India gives an average an-nual profit estimate of pound21m about pound17m in1600 prices


1751ndash1800mdashInikorirsquos British America tradeestimates of pound19545 for 1781ndash1800 impliesannual profits of around pound10m ie doubleOrsquoBrienrsquos profit estimate (approximately pound5min 1600 prices)

It is worth noting that our profit estimateswould be significantly higher prior to 1650 if wealso included British and Dutch trade in Asiangoods passing through Portugal Spain and theLevant (Israel 1989 Brenner 2003)

ReligionmdashFrom Langer (1972) and Stearns(2001) Britain the Czech Republic Denmark

Finland Germany the Netherlands NorwaySweden and Switzerland were majority Protes-tant in 1600 Germany was largely Protestantbut the balance remained unclear until the endof the 1600s The results are robust to codingGermany as Catholic We have also tried analternative specification in which religion iscoded directly as Catholic Muslim Orthodoxor Protestant with essentially identical results

Roman HeritagemdashFrom Langer (1972) thefollowing countries had a Roman heritage Bel-gium Britain France Italy the Netherlands

APPENDIX TABLE 1mdashVARIABLE DEFINITIONS AND SOURCES

Variable Description Source

Log GDP per capita in 1500 1600 1700 1820and 1870

Logarithm of GDP per capita Maddison (2001)

Population in 1000 1200 1300 1400 15001600 1700 1750 1800 and 1850

Total population McEvedy and Jones (1978)

Urban population in 1000 1200 1300 14001500 1600 1700 1750 1800 and 1850

Population living in urban areas Bairoch et al (1988) as described in theAppendix We use Bairoch (1988) forurbanization in Asia and Chandler (1996) forAsian city population

Atlantic and Mediterranean ports City that is on the Atlantic or Mediterranean Bairoch et al (1988) for cities location fromDoring Kindersley (DK) Publishers (1997)

Ratio of Atlantic coastline to area Length of Atlantic coastline divided by land areaBoth assume modern borders Atlantic coastlineincludes the whole coast of Portugal IrelandBelgium the Netherlands and Britain It alsoincludes half the coastline of Spain two-thirdsthe coastline of France half the coastline ofGermany one-quarter the coastline of Denmarkand half the coastline of Norway

Coastline is from Integrated CoastlineManagement (on the Web) Land area is fromthe World Bank World DevelopmentIndicators CD-Rom 1999

Dummy for Atlantic trader Equals one for Britain France the NetherlandsPortugal and Spain

Coded by authors based on composition ofAtlantic trade Acemoglu et al (2002b) fordetails

Dummy for Atlantic port Equals one for a city that was used as an Atlanticport zero otherwise

Bairoch et al (1988) for cities location fromDK Publishing (1997)

Dummy for potential Atlantic port Equals one for a city that is on the Atlantic zerootherwise


Volume of Atlantic trade Average voyages per year equivalent See Appendix Acemoglu et al (2002b) providefull details

Constraint on executive in 1800 1850 19601970 1990 and intervening years

A seven-category scale from 1 to 7 with a higherscore indicating more constraints Score of 1indicates unlimited authority score of 3indicates slight to moderate limitations score of5 indicates substantial limitations score of 7indicates executive parity or subordinationScores of 2 4 and 6 indicate intermediatevalues

Polity IV dataset downloaded from Inter-University Consortium for Political andSocial Research Variable described in Gurr(1997)

Constraint on executive from 1000 to 1800 A seven-category scale from 1 to 7 with a higherscore indicating more constraints Score of 1indicates unlimited authority score of 3indicates slight to moderate limitations score of5 indicates substantial limitations score of 7indicates executive parity or subordinationScores of 2 4 and 6 indicate intermediatevalues

Coded by authors from Langer (1972) seeAppendix for more details

Religion variables Majority religion of city or country Coded by authors from Langer (1972)Roman heritage Coded equal to one for countries that were part of

the Roman Empire and not subsequently part ofthe Ottoman Empire

Coded by authors from Langer (1972)

Wars per year Number of years of war in preceding 50 or 100years Civil wars and colonial wars outsideEurope are excluded

Coded by authors from Kohn (1999)

Latitude Absolute value of the latitude of the countryscaled to take values between 0 and 1 where 0is the equator

Country data from La Porta et al (1999) Citydata from Bairoch et al (1988)


Portugal Spain and Switzerland BulgariaGreece Romania and Yugoslavia had their Ro-man traditions eradicated by a long period ofOttoman rule If they are also coded with Ro-man heritage the effect of this variable is weak-ened further

WarsmdashGeorge Childs Kohn (1999) liststhe dates of every European war from aboutAD 1000 and a brief explanation of partici-pants duration intensity and outcome Wecalculate the average number of years ofwar in a time interval before each date inour dataset for the preceding 100 yearsthrough 1700 and for the preceding 50 yearsfor 1750 1800 and 1850 excluding purelycivil wars and colonial wars outside EuropeAlternative codings such as dropping ldquominorrdquowars does not affect our main results Kohn(1999) does not provide reliable informationon the wars of Finland and Greece during thisperiod so we drop these countries from re-gressions involving the ldquowars per yearrdquovariable

Constraint on ExecutivemdashThis variable iscoded using the method of Polity IV as de-scribed in footnote 32 Our primary source inthis exercise is the historical encyclopedia ofLanger (1972) supplemented with Stearns(2001) Acemoglu et al (2002b) provide moredetails on our coding the full series androbustness checks with some reasonable al-ternatives We also checked our results usingthe three codings of institutions in De Longand Shleifer (1993) which are somewhat dif-ferent from ours for example awarding amuch better score to feudal systems than doescoding based on the Polity criteria Usingtheir measures leads to very similar results tothose reported in the text

REFERENCES

Acemoglu Daron and Robinson James A ldquoPolit-ical Losers as a Barrier to Economic Devel-opmentrdquo American Economic Review 200090(2) pp 126ndash30

Acemoglu Daron and Robinson James A ldquoEco-nomic Backwardness in Political Perspec-tiverdquo National Bureau of EconomicResearch Inc NBER Working Papers No8831 2002

Acemoglu Daron Johnson Simon and RobinsonJames A ldquoReversal of Fortune Geographyand Institutions in the Making of the ModernWorld Income Distributionrdquo Quarterly Jour-nal of Economics 2002a 117(4) pp 1231ndash94

Acemoglu Daron Johnson Simon and RobinsonJames A ldquoThe Rise of Europe AtlanticTrade Institutional Change and EconomicGrowthrdquo National Bureau of Economic Re-search Inc NBER Working Papers No9378 2002b

Anderson Perry Passages from antiquity to feu-dalism London NLB Press 1974

Ashton Robert The crown and the money mar-ket 1603ndash1640 Oxford Oxford UniversityPress 1960

Ashton Robert The city and the court 1603ndash1643 Cambridge Cambridge UniversityPress 1979

Ashton Robert ldquoInsurgency Counter-Insurgencyand Inaction Three Phases in the Role of theCity in the Great Rebellionrdquo in Stephen Por-ter ed London and the Civil War NewYork St Martinrsquos Press 1996

Bairoch Paul Cities and economic develop-ment From the dawn of history to thepresent Chicago University of ChicagoPress 1988

Bairoch Paul Economics and world historyMyths and paradoxes Chicago University ofChicago Press 1993

Bairoch Paul Batou Jean and Chevre Pierre Lapopulation des villes europeenees de 800 a1850 Banque de donnees et analyse som-maire des resultats (The population of Euro-pean cities 800ndash1850 Data bank and shortsummary of results) Centre of InternationalEconomic History Series No 2 Geneva Li-brarie Droz 1988

Bean Richard ldquoWar and the Birth of the NationStaterdquo Journal of Economic History 197333(1) pp 203ndash21

Boxer Charles R Portuguese conquest andcommerce in Southern Asia 1500 ndash1750Reprint London Variorum Reprints1985

Braudel Fernand Civilization and capitalism15thndash18th Century Volume III The perspec-tive of the world Reprint Berkeley Univer-sity of California Press 1992

Braudel Fernand The Mediterranean and theMediterranean world in the age of Philip II


Vol II Reprint Berkeley University of Cal-ifornia Press 1995

Brenner Robert ldquoThe Civil War Politics ofLondonrsquos Merchant Communityrdquo Past andPresent 1973 58 pp 53ndash107

Brenner Robert Merchants and revolutionCommercial change political conflict andLondonrsquos overseas traders 1550ndash1653 Re-print London Verso 2003

Brunton Douglas and Pennington D H Mem-bers of the Long Parliament London Allenand Unwin 1954

Cameron Rondo A concise economic history ofthe world from paleolithic times to thepresent 2nd Ed New York Oxford Univer-sity Press 1993

Carruthers Bruce G City of capital Politics andmarkets in the English financial revolutionPrinceton Princeton University Press 1996

Chandler Tertius Four thousand years of urbangrowth An historical census Lewiston StDavidrsquos University Press 1987

Chaudhuri Kirti N The English East IndiaCompany The study of an early joint-stockcompany 1600ndash1640 London Frank Cass ampCo 1965

Church William F The impact of absolutism onFrance National experience under Rich-elieu Mazarin and Louis XIV New YorkJohn Wiley amp Sons 1969

Cooper J P ldquoSocial and Economic Policiesunder the Commonwealthrdquo in Gerard E Ay-lmer ed The interregnum The quest forsettlement 1646ndash1660 London Macmillan1972

Craven W Frank ldquoThe Earl of Warwick aSpeculator in Piracyrdquo The Hispanic Ameri-can Historical Review 1930 10(4) pp 457ndash79

Davis Ralph The rise of the English shippingindustry in the seventeenth and eighteenthcenturies London Macmillan amp Co Ltd1962

Davis Ralph English overseas trade1500ndash1700 London Macmillan amp Co Ltd1973a

Davis Ralph The rise of the Atlantic economiesIthaca Cornell University Press 1973b

De Long J Bradford and Shleifer AndreildquoPrinces and Merchants European CityGrowth before the Industrial RevolutionrdquoJournal of Law and Economics 1993 36(2)pp 671ndash702

de Vries Jan The economy of Europe in an ageof crisis 1600ndash1750 Cambridge CambridgeUniversity Press 1976

de Vries Jan European urbanization 1500ndash1800 Cambridge MA Harvard UniversityPress 1984

de Vries Jan ldquoConnecting Europe and Asia AQuantitative Analysis of the Cape RouteTrade 1497ndash1795rdquo in Dennis Flynn ArturoGiraldez and Richard von Glahn eds Globalconnections and monetary history 1470ndash1800 London Ashgate Publishers 2003

de Vries Jan and van der Woude Ad The firstmodern economy Success failure and per-severance of the Dutch economy 1500ndash1815 Cambridge Cambridge UniversityPress 1997

Doyle William The Parliament of Bordeaux andthe end of the Old Regime 1771ndash1790 Lon-don Ernest Benn 1974

Doyle William Origins of the French Revolu-tion Oxford Oxford University Press 1988

Dorling Kindersley (DK) Publishers World atlasNew York DK Publishers 1997

Elton Geoffrey R England under the Tudors3rd Ed New York Routledge 1991

Engerman Stanley L ldquoThe Slave Trade andBritish Capital Formation in the EighteenthCentury A Comment on the Williams The-sisrdquo Business History Review 1972 46(4)pp 430ndash43

Engerman Stanley L and OrsquoBrien Patrick KldquoExports and the Growth of the British Econ-omy from the Glorious Revolution to thePeace of Amiensrdquo in Solow Barbara L edSlavery and the rise of the Atlantic systemCambridge Cambridge University Press1991

Farnell J E ldquoThe Navigation Act of 1651 theFirst Dutch War and the London MerchantCommunityrdquo Economic History Review19631964 16 pp 439ndash545

Frank Andre G Dependent accumulationand underdevelopment London Macmil-lan 1978

Fritschy Wantje ldquoA lsquoFinancial RevolutionrsquoReconsidered Public Finance in HollandDuring the Dutch Revolt 1568ndash1648rdquo Eco-nomic History Review 2003 56(1) pp57ndash89

Fritschy Wantje Hart Marjolein and HorlingsEdwin ldquoContinuities and Discontinuities inDutch Fiscal History C 1515ndash1913rdquo Eco-


nomic History Congress Madrid June 21ndash23 2001 httpwwwehnetXIIICongresscdpapers12FritschyHartHorlings86pdf

Gauci Perry The politics of trade The overseasmerchant in state and society 1660ndash1720Oxford Oxford University Press 2001

Grassby Richard ldquoThe Rate of Profit in Seven-teenth-Century Englandrdquo English HistoricalReview 1969 84(333) pp 721ndash51

Grassby Richard The business community ofseventeenth-century England CambridgeCambridge University Press 1995

Graves Michael A R The parliaments ofearly modern Europe New York Long-man 2001

Gurr Ted Robert ldquoPolity II Political Structuresand Regime Change 1800ndash1986rdquo Unpub-lished Paper 1997

Hall John A Powers and liberties The causesand consequences of the rise of the WestOxford Blackwell 1985

Hamilton Earl J ldquoThe Role of Monopoly in theOverseas Expansion and Colonial Trade ofEurope before 1800rdquo American EconomicReview 1948 38(2) pp 33ndash53

Hart Marjolein Jonker Joost and van ZandenJan L A financial history of the NetherlandsCambridge Cambridge University Press1997

Hill Christopher From Reformation to Indus-trial Revolution 1530ndash1780 Baltimore Pen-guin Books 1969

Hintze Otto ldquoMilitary Organization and the Or-ganization of the Staterdquo in Felix Gilbert edThe historical essays of Otto Hintze NewYork Oxford University Press 1975 pp178ndash215

Holmes Geoffrey The making of a great powerLate Stuart and early Georgian Britain NewYork Longman 1993

Horwitz Henry ldquoThe East India Trade thePoliticians and the Constitution 1698 ndash1702rdquo The Journal of British Studies1978 17(2) pp 1ndash18

Inikori Joseph E Africans and the IndustrialRevolution in England A study in interna-tional trade and economic developmentCambridge Cambridge University Press2002

Israel Jonathan I The Dutch Republic and theHispanic world 1606ndash1661 Oxford OxfordUniversity Press 1982

Israel Jonathan I Dutch primacy in world

trade 1585ndash1740 Oxford Clarendon Press1989

Israel Jonathan I The Dutch Republic Its risegreatness and fall 1477ndash1806 Volume I TheOxford history of early modern Europe NewYork Oxford University Press 1995

Jones Eric L The European miracle Environ-ments economies and geopolitics in the his-tory of Europe and Asia New YorkCambridge University Press 1981

Keeler Mary F The Long Parliament 1640ndash1641 A Biographical Study of Its MembersPhiladelphia American Philosophical Soci-ety 1954

Kennedy Paul M The rise and fall of the greatpowers Economic change and military con-flict from 1500 to 2000 New York RandomHouse 1987

Kohn George C Dictionary of war revised edi-tion New York Facts on File Inc 1999

Krusell Per and Rıos-Rull Jose-Victor ldquoVestedInterests in a Theory of Stagnation andGrowthrdquo Review of Economic Studies 199663(2) pp 301ndash30

La Porta Rafael Lopez-de-Silanes FlorencioShleifer Andrei and Vishny Robert W ldquoTheQuality of Governmentrdquo Journal of LawEconomics and Organization 1999 15(1)pp 222ndash79

Landes David S The wealth and poverty ofnations Why some are so rich and some sopoor New York W W Norton amp Co 1998

Lane Frederic C Venetian ships and shipbuild-ing of the Renaissance Baltimore JohnsHopkins University Press 1934

Lang R G ldquoSocial Origins and Social Aspira-tions of Jacobean London Merchantsrdquo Eco-nomic History Review Second Series 197427(1) pp 28ndash47

Langer William L An encyclopedia of worldhistory 5th Ed Boston Houghton MifflinCo 1972

Lefebvre Georges (Translated by R R Palmer)The coming of the French Revolution Prince-ton Princeton University Press 1947

Maddison Angus The world economy A mil-lennial perspective Paris OECD Develop-ment Center 2001

Manning Brian Aristocrats plebeians and rev-olution in England 1640ndash1660 New HavenPluto Press 1996

Marshall Monty G and Jaggers Keith Polity IVProject political regime characteristics and


transitions 1800ndash1999 dataset users man-ual College Park Center for InternationalDevelopment and Conflict ManagementUniversity of Maryland 2000

McEvedy Colin and Jones Richard Atlas ofworld population history New York Factson File Inc 1978

Morrill John S The nature of the EnglishRevolution London Longman PublishingGroup 1993

Neal Larry ldquoHow It All Began The Mone-tary and Financial Architecture of EuropeDuring the First Global Capital Markets1648 ndash 815rdquo Financial History Review2000 7(2) pp 117ndash 40

North Douglass C and Thomas Robert P Therise of the Western world A new economichistory Cambridge Cambridge UniversityPress 1973

North Douglass C Structure and change in eco-nomic history New York W W Norton ampCo 1981

North Douglass C and Weingast Barry R ldquoCon-stitutions and Commitment The Evolution ofInstitutions Governing Public Choice in Sev-enteenth-Century Englandrdquo Journal of Eco-nomic History 1989 49(4) pp 803ndash32

OrsquoBrien Patrick K ldquoEuropean Economic De-velopment The Contribution of the Periph-eryrdquo Economic History Review SecondSeries 1982 35(1) pp 1ndash18

OrsquoRourke Kevin H and Williamson Jeffrey GldquoAfter Columbus Explaining the GlobalTrade Boom 1500ndash1800rdquo Journal of Eco-nomic History 2002 62(2) pp 471ndash56

Olson Mancur The rise and decline of nationsEconomic growth stagflation and economicrigidities New Haven Yale UniversityPress 1982

Parente Stephen L and Prescott Edward CldquoMonopoly Rights A Barrier to RichesrdquoAmerican Economic Review 1999 89(5) pp1216ndash33

Parker Geoffrey The Dutch Revolt New YorkPenguin Books 1977

Parker Geoffrey The military revolution Mili-tary innovation and the rise of the West1500ndash1800 Cambridge Cambridge Univer-sity Press 1988

Parrott David Richelieursquos army War govern-ment and society in France 1624ndash1642Cambridge Cambridge University Press2001

Parry James H The Spanish seaborne empireBerkeley University of California Press1966

Pearl Valerie London and the outbreak of thePuritan revolution City government and na-tional politics 1625ndash43 London OxfordUniversity Press 1961

Phelps Brown Ernest H and Hopkins Sheila VldquoSeven Centuries of Building Wagesrdquo Eco-nomica New Series 1955 22(87) pp 192ndash206

Phillips Carla Rahn ldquoThe Growth and Compo-sition of Trade in the Iberian Empires 1450ndash1750rdquo in James D Tracy ed The rise ofmerchant empires Long-distance trade inthe early modern world 1350ndash1750 Cam-bridge Cambridge University Press 1990pp 34ndash101

Pincus Steven ldquoCivic Republicanism and Polit-ical Economy in an Age of Revolution LawPolitics and Economics in the Revolution of1688ndash89rdquo Unpublished Paper 2002

Rabb Theodore K Enterprise and empire Mer-chant and gentry investment in the expansionof England 1575ndash1630 Cambridge MAHarvard University Press 1967

Rabb Theodore K Jacobean gentleman Sir Ed-win Sandays 1561ndash1629 Princeton Prince-ton University Press 1998

Rajan Raghuram G and Zingales Luigi Savingcapitalism from the capitalists Unleashingthe power of financial markets to createwealth and spread opportunity New YorkCrown Publishing Group 2003

Richardson Roger C The debate on the EnglishRevolution 3rd Ed Manchester Universityof Manchester Press 1998

Rodger N A M The safeguard of the sea Anaval history of Britain 660ndash1649 NewYork W W Norton amp Sons 1997

Rogowski Ronald Commerce and coalitionsHow trade affects Domestic Political Align-ments Princeton Princeton University Press1989

Russell Conrad The causes of the English CivilWar New York Oxford University Press1990

Sacks David H The widening gate Bristol andthe Atlantic economy 1450ndash1700 The NewHistoricism Studies in Cultural Poetics Vol15 Berkeley and Oxford University of Cal-ifornia Press 1991

Scoville Warren C The persecution of


Huguenots and French economic develop-ment 1680 ndash1720 Berkeley University ofCalifornia Press 1960

Sharpe Kevin The personal rule of Charles INew Haven Yale University Press 1992

Stasavage David Public debt and the birth ofthe democratic state France and Great Brit-ain 1688ndash1789 Cambridge New York andMelbourne Cambridge University Press 2003

Stearns Peter N The encyclopedia of world his-tory 6th Ed Boston Houghton Mifflin Com-pany 2001

Steensgaard Niels The Asian trade revolution ofthe seventeenth century Chicago Universityof Chicago Press 1974

Stone Lawrence The causes of the English Rev-olution 1529ndash1642 New York Routledge1972

Thompson I A A ldquoCastile Absolutism Con-stitutionalism and Libertyrdquo in Philip T Hoff-man and Kathryn Norberg eds Fiscal crisisliberty and representative government Stan-ford Stanford University Press 1994

Tilly Charles Coercion capital and Europeanstates AD 990ndash1990 Cambridge BasilBlackwell 1990

Tracy James D A financial revolution in theHabsburg Netherlands Renten and Rent-eniers in the country of Holland 1515ndash1565

Berkeley University of California Press1985

Tracy James D The rise of merchant empiresLong-distance trade in the early modernworld 1350ndash1750 Cambridge CambridgeUniversity Press 1990

Tracy James D ldquoKeeping the Wheels of WarTurning Revenues of the Province of Hol-land 1572ndash1619rdquo in Graham Darby ed Theorigins and development of the Dutch RevoltNew York Routledge 2001

van der Wee Herman ldquoThe Economy as a Fac-tor in the Revolt of the Southern Nether-landsrdquo in Herman van der Wee ed The lowcountries in the early modern world Brook-field Variorum 1993

van Zanden Jan L The rise and decline ofHollandrsquos economy Manchester Universityof Manchester Press 1993

Wallerstein Immanuel M The modern world-system 3 Volumes New York AcademicPress 1974ndash1980

Weber Max The Protestant ethic and the spiritof capitalism London Routledge 1993

Williams Eric Capitalism and slavery ChapelHill University of North Carolina Press1944

World Bank World development indicatorsWashington DC World Bank 1999


Europe and is quite different in nature from theEuropean growth that took place before 1500

Not all societies with access to the Atlanticshow the same pattern of growth however Thedata suggest an important interaction betweenmedieval political institutions and access to theAtlantic the more rapid economic growth tookplace in societies with relatively nonabsolutistinitial institutions most notably in Britain andthe Netherlands In contrast countries where themonarchy was highly absolutist such as Spainand Portugal experienced only limited growthin the subsequent centuries while areas lacking

easy access to the Atlantic even such nonabso-lutist states as Venice and Genoa did not expe-rience any direct or indirect benefits fromAtlantic trade

Figures 1 and 2 illustrate the central thesisof this paper Figure 1 panel A shows thaturbanization in Western Europe grew sig-nificantly faster than in Eastern Europe after15002 Figure 1 panel B shows that these

2 For the purposes of this paper Western Europe is takento be all the countries west of the Elbe ie Austria

FIGURE 1A WESTERN EUROPE EASTERN EUROPE AND ASIA URBANIZATION RATES WEIGHTED BY POPULATION 1300ndash1850

FIGURE 1B ATLANTIC TRADERS WEST EUROPEAN COUNTRIES NOT ATLANTIC TRADERS AND EASTERN EUROPEURBANIZATION RATES WEIGHTED BY POPULATION 1300ndash1850














Wholesample

unweighted

Wholesample

weighted



EasternEurope Asia


Urbanization in 1300 66 99 80 100 41 110(52) (32) (28) (61) (33) (07)

Urbanization in 1400 76 103 85 121 39 111(95) (36) (24) (100) (15) (05)

Urbanization in 1500 83 106 101 114 40 115(76) (34) (53) (68) (18) (07)

Urbanization in 1600 96 117 136 140 44 120(76) (40) (76) (88) (27) (07)

Urbanization in 1700 107 112 145 131 37 116(85) (41) (68) (81) (22) (07)

Urbanization in 1800 141 103 198 169 70 89(91) (49) (79) (75) (33) (14)

GDP per capita in 1500 62754 6083 72146 85073 50694 5750(1593) (1180) (311) (2171) (782) (354)

GDP per capita in 1600 74073 6305 91631 90822 57829 5768(2256) (1442) (1493) (1673) (1123) (353)

GDP per capita in 1700 86212 6222 107921 98082 6360 5742(3484) (2081) (3214) (1282) (1361) (353)

GDP per capita in 1820 98800 6917 132195 109540 7195 5755(3736) (2645) (3487) (1253) (1749) (457)
























A Data













(1) ujt dt j t1600

t WEj dt

t1500

t PATj dt Xjt jt











Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850



Panel1300ndash1850


1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]


0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)




[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]


(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


















(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES












































































































































Wholesample

unweighted

Wholesample

weighted



EasternEurope Asia


Urbanization in 1300 66 99 80 100 41 110(52) (32) (28) (61) (33) (07)

Urbanization in 1400 76 103 85 121 39 111(95) (36) (24) (100) (15) (05)

Urbanization in 1500 83 106 101 114 40 115(76) (34) (53) (68) (18) (07)

Urbanization in 1600 96 117 136 140 44 120(76) (40) (76) (88) (27) (07)

Urbanization in 1700 107 112 145 131 37 116(85) (41) (68) (81) (22) (07)

Urbanization in 1800 141 103 198 169 70 89(91) (49) (79) (75) (33) (14)

GDP per capita in 1500 62754 6083 72146 85073 50694 5750(1593) (1180) (311) (2171) (782) (354)

GDP per capita in 1600 74073 6305 91631 90822 57829 5768(2256) (1442) (1493) (1673) (1123) (353)

GDP per capita in 1700 86212 6222 107921 98082 6360 5742(3484) (2081) (3214) (1282) (1361) (353)

GDP per capita in 1820 98800 6917 132195 109540 7195 5755(3736) (2645) (3487) (1253) (1749) (457)
























A Data













(1) ujt dt j t1600

t WEj dt

t1500

t PATj dt Xjt jt











Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850



Panel1300ndash1850


1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]


0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)




[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]


(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


















(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES
















































































































































A Data













(1) ujt dt j t1600

t WEj dt

t1500

t PATj dt Xjt jt











Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850



Panel1300ndash1850


1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]


0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)




[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]


(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


















(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES








































































































































(1) ujt dt j t1600

t WEj dt

t1500

t PATj dt Xjt jt











Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850



Panel1300ndash1850


1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]


0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)




[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]


(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


















(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES






































































































































Panel1300ndash1850

Panel1000ndash1850

Panel1300ndash1850

Panel1000ndash1850



Panel1300ndash1850


1300ndash1850Panel

1000ndash1850


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[000] [000] [045] [009] [080] [000] [012] [009] [001] [078]


0016 00086 0055 0014 0018 050 038 075(0021) (0019) (0026) (0022) (0016) (068) (065) (087)


0006 0004 00495 00054 00085 021 003 094(0023) (0021) (0028) (0028) (0018) (068) (064) (094)


0032 0022 0071 0032 0024 181 164 201(0021) (0019) (0028) (0026) (0016) (063) (058) (094)


0032 0022 0073 0032 0023 216 199 260(0021) (0018) (0028) (0025) (0015) (062) (057) (094)


0048 0038 0110 0047 0028 330 312 376(0019) (0017) (0028) (0023) (0015) (057) (051) (094)


0085 0076 0115 0084 0043 505 488 467(0018) (0016) (0028) (0022) (0014) (051) (044) (094)




[000] [000] [035] [006] [083] [000] [011] [016] [002] [081]


(00024) (00020) (00034) (00029) (00018) (007) (006) (011)


















(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES














































































































































(2) ujt dt j t1600

t WEj dt

PATj ln ATt Xjt jt


















Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES









































































































































Panel1500ndash1820

Panel1500ndash1870

Panel1500ndash1820

Panel1500ndash1870



Panel1500ndash1820

withoutBritain

Panel1500ndash1820

Panel1500ndash1870


(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)





[044] [005] [092] [023] [017] [001] [089] [097] [058] [031]


014 015 016 014 013 443 446 342(007) (011) (007) (013) (007) (242) (361) (221)


018 019 021 018 014 884 880 632(007) (010) (007) (012) (006) (227) (340) (221)


027 027 018 027 020 1203 1189 806(006) (010) (007) (011) (006) (210) (314) (221)


022 1584(009) (293)




[044] [005] [092] [048] [014] [001] [088] [099] [054] [023]


0069 0040 0047 0069 0051 321 318 222(0016) (0017) (0018) (0028) (0015) (053) (050) (058)










Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES






































































































































Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850



Panel1500 to 1820

controllingfor wars

Panel1500 to 1820


heritage

Panel1500 to 1820


(1) (2) (3) (4) (5) (6) (7) (8)




[067] [042] [049] [009] [024] [091] [015] [085]


0013 0011 0011 0011 0089 0070 0125 0078(0002) (0003) (0003) (0002) (0013) (0017) (0017) (0015)


century00006 0075(0008) (0029)


[089] [000]





[019] [023] [039] [009] [099] [098] [071] [081]


079 076 075 078 278 333 332 296(008) (008) (007) (007) (054) (056) (054) (056)


century00082 0033


year[077] [032]
































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES




























































































































































Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel


Balancedpanel





yearinteractions

Balancedpanel



1300ndash1850with


ports

(1) (2) (3) (4) (5) (6) (7) (8)




[034] [005] [030] [016] [028] [030] [041] [032]

Atlantic port 1500 004 005 0027 0048 0008 0072 003 005(019) (020) (017) (016) (020) (020) (020) (019)

Atlantic port 1600 036 046 041 043 041 036 036 040(016) (020) (014) (016) (017) (017) (016) (016)

Atlantic port 1700 071 062 076 076 0297 047 071 074(014) (020) (013) (016) (017) (017) (015) (015)

Atlantic port 1750 070 071 079 089 026 046 07 072(014) (020) (013) (016) (016) (016) (015) (014)

Atlantic port 1800 079 092 095 110 032 057 0799 084(014) (020) (012) (016) (015) (015) (014) (014)

Atlantic port 1850 109 100 119 123 048 046 109 110(013) (020) (012) (016) (014) (014) (014) (013)


[019]




[023] [004] [023] [010] [031] [033] [030] [020]


017 016 017 016 0065 0078 017 017(002) (002) (0017) (0024) (0019) (0018) (0018) (0017)


[014]





E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES

































































































































E Interpretation











II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES


































































































































II Our Hypothesis

A The Argument










































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES


































































































































































































(3) ujt dt j t1600

t WEj dt

ln ATt PATj

t1500

t Ij1415 dt











Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES






































































































































Panel1300ndash1850

Panel1300ndash1850

Panel1300ndash1850


Panel1300 to 1850

controllingfor wars

Panel1300 to 1850


heritage

Panel1300 to 1850








(1) (2) (3) (4) (5) (6) (7) (8) (9)


[000] [035] [000] [000] [000] [026] [000] [000] [000]


042 2083(047) (2294)


014 1094(052) (2291)


029 6212(048) (2114)


032 8145(046) (2078)


207 7981(044) (1897)


296 7225(041) (1713)


(006) (006) (006) (006) (006) (231)


[000]


0034(020)


[005]










(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES






































































































































(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)


1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850

Panel1000ndash1850



0011 0011 00095 00026 00082 00084 0012 0009(0002) (0002) (00049) (00062) (00020) (00020) (0004) (0005)


[061] [051] [071] [085] [012] [008] [042] [092]


(0004) (0005) (0004) (0004)



1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel

1500ndash1820Panel


Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870

Panel1500ndash1870



0069 0069 0068 0079 0040 0040 0123 0110(0016) (0016) (0028) (0028) (0017) (0017) (0030) (0028)


[040] [031] [0004] [008] [066] [064] [001] [058]


(003) (002) (003) (002)



1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel

1300ndash1850Panel


Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850

Panel1500ndash1850



043 042 0001 0096 035 034 011 015(006) (006) (012) (012) (005) (005) (010) (009)


[027] [014] [0008] [069] [043] [033] [001] [095]


(011) (009) (009) (007)






V Conclusion


























































REFERENCES


































































































































V Conclusion


























































REFERENCES


















































































































































































REFERENCES
































































































































the rise of europe: atlantic trade, institutional change...

Documents