the collective action factors :

Anton Sobolev (CSID HSE)04/04/2012

The Collective Action Factors:Russian Mass Protests 2011-2012

The Collective Action FactorsAnton Sobolev (CSID HSE)04/04/2012

Motivation

2

Why do individuals participate in collective actions?

• (Olson, 1965): small groups VS large groups• The aims of collective actions – specific public goods =>

«free-rider» problem• But how can one explain the participation in large-scale

actions like voting? (Downs, 1957; Tullock, 1968; Fiorina, 1976; Aldrich, 1993; Blais, 2000; Cohen, 2003; Gerber, Rogers, 2009; Bowles, 2012)

• Why do collective actions occur in some cases and but not in the others.

Why collective actions are so rare in Russia?

The Collective Action FactorsAnton Sobolev (CSID HSE)04/04/2012 3

Broad Research Context

«Since we do not yet have a satisfactory theory of when groups are able to solve their collective action problems, our focus will be on the second source of de facto political power» (Acemoglu, Johnson, Robinson, 2006)

• Collective actions determine the success for groups for lobbying their interests, but…

• They also affect the dynamics of political and economic institutions

Collective actions face: 1. strategical cooperation problem 2.a costs of coordination 2.b costs of participation

Collective actionst ⇒ Political power de-factot ⇒ Economic institutionst ⇒ Resource distributiont+1

Resource distributiont ⇒ Political power de-juret ⇒ Political institutionst+1 ⇒ Economic performancet

Politicla institutionst ⇒


Which factors affect costs of coordination and participation in mass actions?

Harsh conditions and low population concentration cause high costs of collective actions, while developing of telecom technologies decreases themPolitical regime matters

Russian mass protests 2011-2012 case

My paper

4


Hypotheses

5

Geography: Population

Concetration, Temperature,

Density of autoroads and railways

Scale of Mass Protests Actions

Falscifications Repressiveness of Political Regime

Telecom Infrastructure


Geography and Collective Actions - 1

6

Country’s border


Geography and Collective Actions - 2

7

Higher population concetration:• Decreases the costs caused by harsh geographical and

weather conditions• Decreases the rate of «free-riders»

(1) (2) (3)


(Tilly, 2003): «the dense population of Paris made it inevitable that in times of trouble large brawls occurred regularly; it did not take much time to see the rise of sizeable mobs such as the crowd that stormed the Bastille on July 14, 1789» (Dowe, 2001): «1848 - Year of Revolutions». Among the 30 most populated cities in Europe in 1800, 11 of the top 15 were shocked by the revolutionary wave 9 , most of them were capital cities of sovereign or vassal states; while none of the next 15 (Compante, Do, 2008): Population Concetration Index («Gridded Population of the World»)

Kuweit VS Saudi ArabiaMovement of capital cities: 9 times since 1960 (Kazakhstan, 1997, Tanzania, 1996)Positive effects (Alesina, Glaeser, 2004)

Evidence

8


Saudi Arabia VS Kuwait: feel the difference

9

Voice and accountabi

lity

Political

stability

Government

effectiveness

Regulatory quality

Rule of law

Control of corruptio

nKuwait-1996 -0.47 0.01 0.35 -0.04 0.74 0.61Saudi Arabia-1996 -1.62 -0.52 -0.34 -0.38 0.45 -0.42Kuwait-2007 -0.46 0.4 0.2 0.29 0.69 0.49Saudi Arabia-2007 -1.59 -0.59 -0.18 -0.1 0.27 -0.1


(Treivish, 2003): Even in the most populous regions poplation density remains at least 2 times lower than in the nearest western post-communist countries (Ekiert, Hanson, 2003; Getachew, Lankina, 2006): The succsefulness of political and economic reforms in post-soviet world well-defined territorial distribution(Turovsky , 2005): distances between russian cities make them «highly fragmented, isolated, immersed in their own problems» (Gaddy, Hill, 1999; Mikhailova, 2005): Low temperature increases transport and living costs of households, decreases personal welfare

Russian case: Geographical Variation

10


(Miner, 2011): Internet technologies decreases the costs of cooperation. (The rise of opposition voters’ turnout in internet-developped regeions in Malasya election, 2008)(Tucker, 2007): Falscifications: difference between median voter’s ideal point and the official policy(Acemoglu, Robinson, 2006; 2004; Oleinik, 2010; Lukes, 1971; Flores, Smith, 2010; De Mesquita, Smith, 2010) : Repressiveness of Political Regime(King et al, 2012): Great Chinese Firewall Case

Factors of Participations: IT and Politics

11


Cyclone Nargis (Burma, 2008):

138 000 dead«Dead people cannot protest»

Are You Ready to Repress?

12


(Robertson 2010): «Despite a broad economic and social crisis, however, protests was concentrated in a small number of very highly mobilized regions»(Smyth, Sobolev, Soboleva 2013): Positive impact of Twitter and Facebook on probability to participate in pro/anti-government rally(Tucker 2007; Putilin 2012): Protests are the immediate result of perceptions of extensive electoral fraud(Enikolopov, Makar’in, Petrova, Polishcuk 2013): An important incentive for people go for protests is that other people are participating

Russian Mass Protests

13


Dataset: 440 mass actions / 309 antigovernment ralliesIntegrum regional newspapers databaseMass actions occurred in 75 regionsEstimated number of participants in the most large-scale antigovernment rally in the region, 4.11.2011-30.05.2012 (Organizators, UVD, mean)Control for population sizeControl for territory

Data – 1: Dependent Variable

14


Poplation concetration index (Herfindahl–Hirschman Index formula for the urban population):

PCI = S12+ S2

2 + S32+…+ Sn

2,Control for rural populationControl for size of the regional area

Data – 2: Independent variables

15


Geography: mean temperature of January, distance from Moscow (Rosstat)Telecom: share of households internet users, number of mobile phone per capita, number of PC per caputa, number of fixed phones per capita Falscifications: share of fraud in elections (Shpilkin, 2011)Repressiveness of political regime: Democracy score index by N. Petrov-A.Titkov, Carnegie (2004)

Data – 2: Independent Variables

16


Numbers: Organizators VS Uvd

17

Kernel Density Estimation of Logged Number of Participants of the Most Crowded Protest Meeting in the Region


Geography. Population Concentration

18


Temperature

19


Telecom Infrastructure

20

(10) (11) (12) (13) (15) (16) (17) (18) VARIABLES N of participants per ca (ORG), % share phone 0.00101

(0.000661) share mobile

-0.00111

(0.00165) share pc

0.00115

(0.000692) share internet

0.00156**

(0.000688) pc

0.00205**

(0.000931) mobile

2.65e-05

(0.000261) density phone

0.00351***

(0.00110) mobile

abonents 0.000226

(0.000398)

PCI rural share Controls for

Msk and Spb Yes Yes Yes Yes Yes Yes Yes Yes

Observations 81 81 81 81 82 82 82 82 R-squared 0.585 0.580 0.587 0.590 0.596 0.577 0.613 0.579

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1


Economics

21

(1) (2) (4) (5) (6) (7) (8) VARIABLES N of participants per ca (ORG), % GRP 8.81e-07

(8.86e-07) unemployment

-0.00515*

(0.00309) cost of living

9.83e-06

(1.45e-05) wage

2.10e-05

(1.79e-05) GINI

2.499

(1.512) reg_heductoempd2010

0.0121

(0.00933) auto2009

0.000683

(0.000474) Observations 79 82 82 79 81 82 82 R-squared 0.095 0.049 0.010 0.092 0.237 0.165 0.054



Economics with controls

22

(9) (10) (12) (13) (14) (15) (16) VARIABLES N of participants per ca (ORG), % GRP -1.87e-07*

(9.64e-08) unemployment

-0.000640

(0.000888) cost of living

-1.37e-05***

(4.71e-06) wage

-7.35e-06**

(3.40e-06) GINI

0.252

(0.328) heductoempd

-0.00284

(0.00250) auto

-2.15e-05

(0.000198)

PCI 0.213*** 0.220*** 0.232*** 0.221*** 0.222*** 0.236*** 0.221***

(0.0531) (0.0542) (0.0514) (0.0526) (0.0534) (0.0517) (0.0537)

rural share -0.00328*** -0.00260*** -0.00332*** -0.00363*** -0.00263*** -0.00291*** -0.00281***

(0.000706) (0.000771) (0.000629) (0.000733) (0.000657) (0.000671) (0.000703)

Msk 0.864*** 0.844*** 0.848*** 0.860*** 0.814*** 0.892*** 0.844***

(0.0523) (0.0510) (0.0449) (0.0490) (0.0697) (0.0881) (0.0521)

Observations 79 82 82 79 81 82 82 R-squared 0.776 0.767 0.783 0.780 0.768 0.773 0.767



Politics: Fraud VS Repressivness

23


Politics. Repressiveness

24


Final model

25

(1) (2) (3) (4) (5) (6) (7) (8) VARIABLES N of participants per ca (ORG), % PCI 0.238*** 0.229*** 0.249*** 0.252*** 0.238*** 0.217*** 0.249*** 0.237***

(0.0558) (0.0528) (0.0543) (0.0511) (0.0558) (0.0584) (0.0543) (0.0550)

rural share -0.000941 -0.000671 -0.00141* -0.00146** -0.000941 -0.000865 -0.00141* -0.00152*

(0.00101) (0.000892) (0.000801) (0.000723) (0.00101) (0.00102) (0.000801) (0.000797)

reg_tempjan 0.00173* -3.60e-05 0.00190* 0.000558 0.00173* 0.000861 0.00190* 0.00130

(0.00102) (0.00123) (0.00111) (0.00126) (0.00102) (0.00117) (0.00111) (0.00118)

Democracy ind 0.00476*** 0.00443*** 0.00472*** 0.00428*** 0.00476*** 0.00450*** 0.00472*** 0.00435***

(0.00128) (0.00120) (0.00132) (0.00127) (0.00128) (0.00128) (0.00132) (0.00136)

density_phone 0.00178 0.00350**

0.00178 0.00284*

(0.00138) (0.00157)

(0.00138) (0.00155)

cost of living

-1.77e-05***

-1.57e-05**

(5.84e-06)

(6.43e-06)

share_internet

0.000745 0.00167*

0.000745 0.00134

(0.000813) (0.000873)

(0.000813) (0.000949)

GRP

-2.75e-07**

-2.17e-07*

(1.12e-07)

(1.25e-07)

Msk Yes Yes Yes Yes Yes Yes Yes Yes Observations 80 80 80 80 80 78 80 78 R-squared 0.808 0.825 0.806 0.819 0.808 0.815 0.806 0.812



Robust?

26

(17) (18) (19) (20) (21) (22) (23) (24) VARIABLES N of participants per ca (ORG), % PCI 0.248*** 0.240*** 0.259*** 0.266*** 0.248*** 0.228*** 0.259*** 0.246***

(0.0575) (0.0541) (0.0556) (0.0510) (0.0575) (0.0605) (0.0556) (0.0562)

rural share -0.000447 -0.000149 -0.000243 2.01e-05 -0.000447 -0.000383 -0.000243 -0.000240

(0.000981) (0.000834) (0.000955) (0.000853) (0.000981) (0.000986) (0.000955) (0.000969)

reg_tempjan 0.00150 -0.000323 0.00181* 3.83e-05 0.00150 0.000663 0.00181* 0.00105

(0.00101) (0.00120) (0.00107) (0.00120) (0.00101) (0.00115) (0.00107) (0.00114)

Democracy ind 0.00457*** 0.00422*** 0.00413*** 0.00341** 0.00457*** 0.00433*** 0.00413*** 0.00368**

(0.00132) (0.00121) (0.00142) (0.00132) (0.00132) (0.00132) (0.00142) (0.00147)

density_phone 0.00130 0.00304*

0.00130 0.00236

(0.00146) (0.00163)

(0.00146) (0.00162)

cost of living

-1.82e-05***

-2.03e-05***

(5.97e-06)

(6.76e-06) share_internet

0.00140 0.00277***

0.00140 0.00219**

(0.000906) (0.00101)

(0.000906) (0.00106)

GRP

-2.70e-07**

-2.89e-07**

(1.16e-07)

(1.20e-07)

republics Yes Yes Yes Yes Yes Yes Yes Yes Msk Yes Yes Yes Yes Yes Yes Yes Yes Observations 80 80 80 80 80 78 80 78 R-squared 0.818 0.835 0.820 0.841 0.818 0.824 0.820 0.829



In regions with lower density concentration and harsh conditions collective actions have less scale character. This correlation is robust even if we control for all political, economic and telecom factors. Factors which affect the costs of participation overlap with the problems of strategic cooperation. The growth of collective action potention is important factor for incumbent behaviorBut the political elite’s response is also important: repertoire of instruments to manage protests action is great (from oppositional leaders’ presecution to capital movement)

Conclusions

27


Coincidence?

28

the collective action factors :

Documents

collective actionfactors

collective action problems

mass actions

largescale actions

costs of coordination

higher population concetration

population concetration

b costs of participation