Domestic resource mobilization. Infrastructure

Development Finance Network (DeFiNe)

Annual Meeting

Paris, 10-12 October 2010

Setting the scene: Infrastructure patterns in emerging markets

Christian Daude and Ángel Melguizo

Americas Desk

OECD Development Centre

2

• Infrastructures are key for potential growth, development and stabilization policies (e.g. G20 agenda)

Growth and inequality gaps Asia-Latin America explained by infrastructure gaps – less spending, lower quality (Calderón and Servén, 2004b)

• Emerging economies: significant infrastructure gaps

• Latin America lags behind Asia and emerging Europe

• Significant differences across infrastructure types (basic, transport, energy and telecommunications) and countries

• Fundamentals-observed levels (Balmaseda, Daude, Melguizo and Taft, 2010)

• Policy response

• Building better institutions (quality of bureaucracy, fiscal position)

• Improving regulation (in particular around public-private financing)

Main messages

3

Developed

Eastern Europe

Asia LatAm

10

11

12

13

14

0 1 2 3 4 5

Log GDP pc

Log lin

es/p

op

Setting priorities: infrastructure levels

Per capita telephone lines and Income level, 2007

Income and population matter – but so do demographic (age profile), social (urbanization) and economic (sector mix) variables.

Source: Balmaseda et al. (2010)

LatAm

Asia

Eastern Europe

Developed

3

4

5

6

7

8

9

10

20 40 60 80 100 120

Urbanization ratioLog K

ws/p

op

Kilowatts pc and Urbanization rate, 2006

Source: Balmaseda et al. (2010)

4

Empirical model (Balmaseda et al., 2010)

Explanatory variables - Per capita income (level and squared) - Socio- demographics (urbanization, density) - Productive structure (services and industry vs. agriculture)

Predicted infrastructure patterns (Km/area, KW pc, pc lines)

Observed levels (Km/area, KW pc, pc lines)

‘Degree of achievement’ (Observed levels/ Patterns)

Predicted infrastructure patterns (for country i, in time t) can be compared to actual levels, to estimate gaps and identify priorities.

5

Results (observed vs. predicted): Priorities

Paved roads

Asian and Lat. Am. challenges concentrated in Transport and Energy infrastructure. In LAC, even the regional leader is below predicted levels.

0%

40%

80%

120%

160%

200%

240%

280%

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

LatAm MAX-min LatAm

Asia

Eastern Europe

Observed / Predicted (%)

0%

20%

40%

60%

80%

100%

120%

140%

160%

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

LatAm MAX-MIN LatAm

Asia

Eastern Europe

Electricity Capacity Generation

Source: Balmaseda et al. (2010) Source: Balmaseda et al. (2010)

6

Results (observed vs. predicted): Priorities

Telephone lines

The situation in telecommunication and basic infrastructure is more balanced. Some good practices may stem from LAC.

0%

50%

100%

150%

200%

250%

300%

350%

400%

450%

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

LatAm

Asia

Eastern Europe

LatAm MAX-min

Access to improved water

70%

90%

110%

130%

150%

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

LatAm

Asia

Eastern Europe

LatAm MAX-min

Observed / Predicted (%)

Source: Balmaseda et al. (2010) Source: Balmaseda et al. (2010)

7

Results (gaps and fiscal balances): Domestic financing

Lower public debt ratios are correlated with lower infrastructure gaps • Fiscal consolidation have been traditionally based on investment cuts (Calderón and Serven, 2004a, Martner and Tromben, 2005 for LAC) • Public borrowing costs reflect (perceptions of) debt sustainability

-20

24

6

e(

ecg

| X

)-.5 0 .5 1 1.5

e( debtgdp | X )

coef = -.70912416, (robust) se = .09684161, t = -7.32

-20

24

6

e(

rail

| X

)

-.5 0 .5 1 1.5e( debtgdp | X )

coef = -.52702326, (robust) se = .18359908, t = -2.87

Electricity Railways

Source: Balmaseda et al. (2010) Source: Balmaseda et al. (2010)

8

Results (gaps and bureaucracy): Public administration

-20

24

6

e(

ecg

| X

)-3 -2 -1 0 1

e( bqual | X )

coef = .57518505, (robust) se = .03843958, t = 14.96

-20

24

e(

paved

| X

)

-3 -2 -1 0 1e( bqual | X )

coef = .45977591, (robust) se = .03161675, t = 14.54

Electricity Paved roads

Better institutions (quality of bureaucracy) are correlated with lower infrastructure gaps

• Need to have domestic resources and management capacity • Governance, control of corruption, adequate regulation

Source: Balmaseda et al. (2010) Source: Balmaseda et al. (2010)

Domestic resource mobilization. Infrastructure

Development Finance Network (DeFiNe)

Annual Meeting

Paris, 10-12 October 2010

Setting the scene: Infrastructure patterns in emerging markets

Christian Daude and Ángel Melguizo

Americas Desk

OECD Development Centre

www.oecd.org/dev/americas

10

Annex: Database

LatAm AsiaEastern

EuropeOther Asia Europe

North

America

Arg China Bul Isr Kor Austria US

Bra India Cro Tur HK Bel Can

Chile Indo Cz Rus Sing Fra

Col Mal Hun Jor Jap Ger

Mex Phi Lit South Af. Aud Gre

Ven Tha Est Egy NZ Ire

Peru Vietnam Pol Mauritania Ita

Costa Rica Rom Mor Net

Dom. Republic Slovak Por

Slovenia Spa

Ukr OK

Lux

Swi

Den

Fin

Nor

Swe

Ice

Emerging Economies Developed

11

Annex: Database

Number of

countriesSample

Infrastructure Stocks

Telephone lines (mobile and fixed) Number of lines 61 80-07

Electricity Generating Capacity Kilowatts 60 80-06

Paved Roads Kilometers 61 80-04

Rail-lines Kilometers 60 80-06

Improved water source (% of population with access) % 51 90-06

Sanitation (% of population with access) % 46 90-06

Number of Personal Computers In million 61 93-06

Number of Internet users In million 61 93-06

Other variables

Population In million 61 80-07

Gross Domestic Product In constant 2005 PPPs 61 80-07

Land area In squared kilometers 61 80-07

Urbanization ratio% of urban population over

total61 80-07

Share of Industry, Services, Agriculture on GVA % 61 80-07

Canning (1998) and Canning and

Farahani (2007), extended with WB

World Development Indicators . For

electricity, 2006 from the United

Nations’ Energy Statistics.

World Development Indicators

World Development Indicators

Variable Units Source

Coverage

12

Annex: Results

Regressions at a glance

Note: Panel estimation. Telecoms and Energy regressions include temporal fixed effects. Bold, significant at 5 per cent

Source: Balmaseda et al. (2010)

Empirics

Energy

Water Sanitation Electricity Roads Railways Telephones PC Internet

Income pc + + + + + + + +

Income pc2 - - - - - -Urbanization + + + +

Density + + + + + + + +

Services + + + + + + +

Industry + + + + + +

Basic Transport Telecommunications

13

Annex: Results

Results at a glance

Observed / Pattern

(%, weighted average)

Basic Energy Transport Telecos Total

LatAm 115% 55% 26% 103% 81%

Emerging Asia 79% 53% 55% 204% 117%

Eastern Europe 100% 125% 247% 161% 163%

Developed 102% 143% 185% 75% 118%

Source: Balmaseda et al. (2010)