0. 1

1. 1. Introduction
The relationship between democracy and economic growth has received considerable attention in recent years
But the debate on whether democracy promotes or hinders growth is centuries old (Hobbes 1651; Harrington 1656)
These concepts have changed in the last three centuries
Evidence on the causal relationship remain quite inconclusive
We use a cointegrated Vector Autoregressive approach to investigate the impact of democracy on economic growth
In the context of Bangladesh 1

2. 2. Motivation
Bangladesh provides a unique case to study the democracy‐economic growth relationship
There have been periods of both democracy and autocracy (in the form of military rule) in the last three decades
At the same time Bangladesh has experienced varying economic performance
Helps us analyze the reciprocal links between GDP growth and democracy on Bangladesh
By estimating a VAR and SVAR over 2

3. 3. Three Schools of Thoughts
Conflict view: Democracies impede economic growth
Complementary view: Democratic regimes spur growth; political pluralism is critical for economic pluralism
Skeptical view: Democracy is for the rich? 3

4. 3.1. Three Schools of Thoughts
Fear that democracy hindered economic development
Galenson (1959), Huntington (1968), O’Donnell (1973)
Rise of unions > pressures on wages/ consumption/ inflation
Shift from investment to consumption
Democratic regimes more vulnerable to public pressures
Stronger governments can take difficult decisions in long-term national interest (the ‘authoritarian advantage’) 4

5. 3.1. Three Schools of Thoughts
Counter arguments
North (1990), Barro (1990), Olson (1991), Sen (1994)
Democracies better allocate resources to productive uses
Some government intervention in economy is optimal for growth
Dictatorships are less efficient than markets
Sen: “no democracy ever experienced a famine” – press & opposition 5

6. 3.2. Is There a Trade-off?
Do developing countries have to choose either higher per capita income/less democracy
The ‘China/Singapore model’
Democratization and poorer economies
The ‘India’ route? 6

7. 4. Economic Performance of Bangladesh
Socioeconomic indicators have improved over time
Particularly after the advent of ‘new democracy’ in 1990 7

8. 5. Theoretical Framework
Nearly all theoretically informed empirical studies of economic growth begin with the neoclassical model
Originally proposed by Solow (1956) and extended by Mankiw, Romer, and Weil (1992)
We follow a similar path
Controlling for political regime characteristics, physical investment, human capital, investment climate, and institutional capacity 8

9. 6. Econometric Framework
We utilize a VAR approach
𝑦 𝑡 = 𝛼+ 𝛽𝑑𝑒𝑚𝑜𝑐𝑟𝑎𝑐𝑦 𝑡 + 𝜆𝒁 𝑡 + 𝜀 𝑡
yt: real GDP per capita
democracy: Polity IV regarding the quality of democracy
Zt: vector of control variables; gross capital formation, education expenditure/GDP, government expenditure/GDP
Rule of law and control of corruption; World Governance Indicators 9

10. 7. Data and Variables
Economic growth: log difference of real GDP per capita
Political regime: Democracy and autocracy scores and their difference (polity)  
Population growth:, a flow measure for growth of the labor force, estimated by the annual log difference of population
Investment: log of gross capital formation
Government expenditure: log of the share of real final government expenditure as a share of GDP 10

11. 7.1. Data and Variables – Polity IV
Examines concomitant qualities of democratic and autocratic authority in governing institutions
Incorporates component measures such as key qualities of;
Executive recruitment, constraints on executive authority,
Political competition
Changes in the institutionalized qualities of the authority
Only institutions of the government and political groups within the scope of the authority are included 11

12. 8. Stationarity and Cointegration
Unit root tests
All the explanatory  variables except population growth are non-stationary at their levels
But stationary when their first differences were taken  
None of the series are integrated beyond (1)
Bound test for cointegration can be applied
Bounds tests for cointegration
Reject null hypothesis of no integration
Evidence of a long-run relationship among the regressors
Including democracy and autocracy, and GDP growth 12 1

13. 9. Results Variable Coefficients & S.E. Lagged GDP 0.946* (0.034)*
(0.004)
Government spending
0.629***
(0.119)
0.502***
(0.205)
Gross capital formation
0.0012**
(0.001)
0.0105**
Education expenditure
0.214***
(0.02)
0.220***
(0.065)
Population growth
0.0404***
(0.009)
0.0391***
(0.006)
Democracy **
(0.003) -
Polity2
0.007
(0.693)
Constant
-0.143
(1.604)
-0.129
(1.802)
Observations 30
Including democracy and autocracy, and GDP growth 13 1

14. 9.1. Results
Democracy, as practiced in Bangladesh – has no significant impact in promoting economic growth
Effect of a change in political regime (whether democratic or not) seems to be negatively significant after one time‐period
Even when both the degree and the duration of democracy are considered, there is no impact on economic growth
After one year, unrestricted executive power, democratically elected or not, starts to become ‘autocratic’ in practice 14 1

15. 9.2. Results Rule of law has positive and significant impact on growth
Control of corruption seems to have negative effect
Democracy needs effective rule of law to have a positive impact
There is evidence of export‐led growth in Bangladesh
Government spending and education expenditure seems to have a significant impact on GDP per capita
The coefficient for capital formation is rather small; the impact of investment has not been as high as expected 15 1

16. 10. Conclusion Could be as a result of;
Democracy, as practiced in Bangladesh, seems to have had no significant impact on economic growth of the country
Could be as a result of;
Inefficient economic policy design
Lack of consensus on economic policy and policy continuation, inadequate democratic decision making process
Pressures from lobbying groups, weak institutions, and lack of institutional reforms required to promote economic growth 16 1

17. 10.1. Conclusion
It is often argued that democracy in Bangladesh is essentially limited to casting votes once in five years
Could be another reason why democracy has no significant impact on economic growth in Bangladesh
It really is about institutions, isn’t it? 17 1