the power of institutions: the importance of effective ...presenter.pdfthe power of institutions:...

25
The power of institutions: The importance of effective governments for electricity use in developing countries Crawford School PhD conference 22 November, 2016 Rohan Best

Upload: trantram

Post on 11-Apr-2018

219 views

Category:

Documents


2 download

TRANSCRIPT

The power of institutions: The importance of

effective governments for electricity use in

developing countries

Crawford School PhD conference

22 November, 2016

Rohan Best

2

Electricity consumption per capita less

than 20% of developing country average

3

Electricity is important: SDG goal 7

4

Electricity is important (continued)

• Economic impacts

– Electricity is an important factor in production

• Social impacts

– Students can study at night (for example)

• Health impacts from using other energy sources

– Use of solid fuel instead of electricity can lead to

indoor pollution

• Environmental impacts from other energy sources:

– Deforestation

– Climate change

5

Governance and electricity

• External benefits and costs from electricity use

• Private markets alone may not achieve desired

outcomes

• Important role for public sector

• But which aspect of governance is the most

important?

6

Research question

Which governance attribute is the most

important for the electricity sector in

developing countries?

7

Research answer

Effectiveness of governments is more

important than other attributes such as

political stability or control of corruption

8

Government effectiveness and

electricity consumption, 2012

9

Government effectiveness and

electricity access, 2012

Potential contributors to electricity sector

development

10Three Gorges Dam: Source: www.absolutechinatours.com

11

Potential contributors to electricity

sector development (continued)• Geographic factors

– Natural endowments

– Temperature

– Population density

• Demographic factors

– Rural population (% of total population)

• Economic factors:

– GDP per capita

– Electricity price

• Governance attributes

– Degree of democracy

– Government effectiveness

12

National-level data

• World Development Indicators (World Bank)

– 135 low and middle-income countries

– Levels in 2012

• United Nations

• World Economic Forum

• Worldwide Governance Indicators

• State Fragility Index, Polity IV

• Climate Research Unit

• Clio Infra Databases

13

Method: Cross-sectional regression

14

Electricity dependent variables• Six distinct electricity variables:

– Electricity capacity

– Household electricity consumption

– Final electricity consumption

– Electricity access (% of population)

– Transmission and distribution losses (% of output)

– Electricity quality – reliability of supply

• Levels

• Changes

– Initial electricity use is also a control

15

Government effectiveness definition

• Government effectiveness (WGI)

– Quality of public services and quality of policy

formulation and implementation

– Index values distributed over standard normal

– Combines views of a large number of firms, citizens,

experts through 48 indicators

– 2 of 48 indicators refer to electricity:

• Coverage of electricity grid

• How problematic is electricity for growth of business

– State effectiveness (SFI) is a similar variable

• Does not mention electricity

16

Results• Significant variables supporting electricity use:

– Government effectiveness

– Log GDP per capita

• Significant variables detracting from electricity

use:

– Temperature

– Rural population percentage

– Africa/Middle East binary variable

17

Results: OLSLn

electricity

capacity

per capita

Ln household

electricity

cons. per

capita

Ln final

electricity

cons. p.c.

Electricity

accessElectricity

losses

Electricity

quality

Gov. effectiveness

1996

0.310* 0.318** 0.533*** -0.007 -0.050** 0.541**

Ln GDP pc, 1996 0.604*** 0.799*** 0.673*** 0.157*** -0.014 0.351*

Rural population,

1996

-0.019*** -0.012** -0.016*** -0.005*** 0.001 -0.009

Population

density

-0.080 0.340 0.106 0.297* -0.064 -1.023

Temperature -0.050*** -0.040*** -0.053*** -0.008*** -0.001 -0.011

Water resources

p.c.

0.003** 0.001 0.001 -0.000 0.000 -0.004**

Electricity price -0.017 -0.026*** -0.015 -0.002 0.004* 0.019

Democracy /

autocracy

-0.023 -0.010 -0.021* -0.003 0.001 -0.052**

Africa / Middle

East

-0.927*** -0.619*** -0.695*** -0.244*** 0.061** -0.831**

R squared 0.804 0.831 0.843 0.812 0.213 0.424

18

Possible statistical issues• Potential sources of endogeneity

– Government effectiveness

• Reverse causation

• Omitted variables e.g. technology, security,

policies

– Economic variables

19

Addressing possible issues

• Instrumental variable for government effectiveness

– To instrument for government effectiveness in 2012, I

use log life expectancy from 1950-1955

• Historical economic and demographic variables

• Drop endogenous variables without deep lag

• Changes in dependent variables

20

Results: IV, historicalLn

electricity

capacity

per capita

Ln household

electricity

cons. per

capita

Ln final

electricity

cons. p.c.

Electricity

accessElectricity

losses

Electricity

quality

Gov. effectiveness

IV

1.776*** 1.974*** 1.890*** 0.224*** -0.189** 0.758

Ln GDP pc, 1960 0.330 0.400 0.467 0.066 0.087 0.788**

Rural population,

1960

-0.032*** -0.019 -0.020** -0.010*** 0.002 0.001

Population

density

-0.950 -0.786 -0.991 0.137 -0.001 -1.689**

Temperature -0.034** -0.046*** -0.046*** -0.007** 0.002 -0.010

Water resources

p.c.

-0.000 -0.000 -0.005 -0.001 -0.001 -0.011

Africa / Middle

East

-0.323 -0.110 -0.217 -0.155*** -0.034 -0.537

R squared 0.579 0.508 0.625 0.680 0.066 0.540

Second-stage F

stat25.50*** 17.95*** 30.49*** 41.33*** 1.12 10.78***

F stat (excluded) 11.859 11.679 11.859 11.859 4.052 6.933

Partial R squared 0.166 0.173 0.166 0.166 0.086 0.161

21

Results: ChangesEl. Capacity

p.c. Δ % p.a.

(1996-2012)

H/h El.

cons. p.c.

Δ % p.a.

(1996-2012)

Final El.

cons. p.c. Δ

% p.a.

(1996-2012)

Electricity

access Δ

(2000-

2012)

Electricity

losses Δ

(1996-

2012)

Electricity

quality Δ

(2006-

2012)

Government

effectiveness (1996)

0.011** 0.016*** 0.019*** 0.021* -0.038 -0.020

Ln GDP pc (1996) 0.021*** 0.017*** 0.021*** 0.006 -0.023 0.489***

GDPpc growth 96-12 0.800*** 0.792*** 0.883*** 0.115 -1.708** 17.885***

Ln electricity

capacity p.c. (1996)

-0.019***

Ln H/h electricity

cons. p.c. (1996)

-0.020***

Ln final electricity

cons. p.c. (1996)

-0.022***

Electricity access

(2000)

-0.099***

Electricity losses

(1996)

-0.866***

Electricity quality

(2006)

-0.438***

R squared 0.347 0.411 0.504 0.203 0.552 0.354

22

Implications• Important role for public sector

• Institutions are important for electricity sector

development, in addition to being important for

economic growth more broadly

• In particular, the important role of government

effectiveness is confirmed and emphasized

• Policy focusing on improving government

effectiveness is an indirect way to contribute to

electricity sector development

Thanks

23

Appendix

24

Source for title slide picture:

https://www.mccombs.utexas.edu/Events/Austin-Electricity-Conference

25

Results compared with other literature

• Onyeji et al. (2012) also finds that effective governments

are associated with higher electricity access for a sample

of 60 developing countries

– This presentation uses a sample of up to 135

countries, multiple electricity variables including

consumption and capacity, and controls for important

variables relating to geography and economics

• Ahlborg et al. (2015) find that institutional quality is

positively associated with household electricity

consumption per capita for a sample of African countries.