Development Economics ECON 4915 Lecture 11 Andreas Kotsadam

  • Slides: 66
Download presentation
Development Economics ECON 4915 Lecture 11 Andreas Kotsadam

Development Economics ECON 4915 Lecture 11 Andreas Kotsadam

Outline • Question from you. • Possible exam question and a recap. • Last

Outline • Question from you. • Possible exam question and a recap. • Last year’s exam is posted on the web.

Outline continued • Big question today: • Why are some countries poor and other

Outline continued • Big question today: • Why are some countries poor and other countries rich? Ø Introduction. Ø Democracy and development. Ø Acemoglu et al. (2001). Ø Glaeser et al. (2004). Ø Bruhn and Gallego (2012) and Heldring and Robinson (2012).

Questions from you • What is the relationship between democracy and economic development (today).

Questions from you • What is the relationship between democracy and economic development (today). • There is a lot of time for recap during the last two lectures, tell me if there is something you want covered!

Possible exam question • Olken and Pande (2011) discuss various ways of measuring coruption.

Possible exam question • Olken and Pande (2011) discuss various ways of measuring coruption. Go through these different types of measures, provide examples, and explain the measures carefully.

Why are some countries poor and other countries rich? Blog: http: //whynationsfail. com/

Why are some countries poor and other countries rich? Blog: http: //whynationsfail. com/

History of economic growth Adam Smith 1776, Malthus 1798

History of economic growth Adam Smith 1776, Malthus 1798

Why are some countries poor and other countries rich? • The geography hypothesis. Three

Why are some countries poor and other countries rich? • The geography hypothesis. Three common variants: Ø Climate may affect productivity directly. Ø The burden of infectious disease is higher in the tropics thanin the temperate zones. Ø Geography may determine the technology available to a society.

 • National geographic video here: http: //www. youtube. com/watch? v=bgnm. TY_r. GQ

• National geographic video here: http: //www. youtube. com/watch? v=bgnm. TY_r. GQ

Arguments (1): • The first step towards civilization is the move from hunter-gatherer to

Arguments (1): • The first step towards civilization is the move from hunter-gatherer to agriculture, with the domestication and farming of wild crops and animals. • Agricultural production leads to food surpluses, which supports sedentary societies, rapid population growth, and specialization of labor.

Arguments (2): • Large societies tend to develop ruling classes and supporting bureaucracies, which

Arguments (2): • Large societies tend to develop ruling classes and supporting bureaucracies, which may lead in turn to the organization of nation states and empires. • Eurasia gained an early advantage due to the greater availability of suitable plant and animal species for domestication. This in turn is due to Eurasia's large landmass and long east-west distance.

Arguments against the hypothesis • Differences across neighboring countries (North/South Korea, East/West Germany, Mexico/US).

Arguments against the hypothesis • Differences across neighboring countries (North/South Korea, East/West Germany, Mexico/US). • The tropics in the Americas were much richer than the temperate zones at the time of colonialization. • Hence, the ”obvious fact” of tropical poverty is neither obvious nor a fact.

Arguments against the hypothesis • Disease is largely a consequence of governments being unable

Arguments against the hypothesis • Disease is largely a consequence of governments being unable or unwilling to take the necessary measures to eradicate it (UK very unhealthy, US and Australia eliminated malaria). • Diamond cannot explain the inequality we see within continents (Norway vs India).

More arguments against • It cannot explain the reversal of fortune in Latin America,

More arguments against • It cannot explain the reversal of fortune in Latin America, nor that the Middle East once led the world, that the first towns developed in modern Iraq, that iron was first melted in Turkey. • Cannot explain why many nations stagnate for long periods and then start growing really quick.

Why are some countries poor and other countries rich? • The culture hypothesis. •

Why are some countries poor and other countries rich? • The culture hypothesis. • Weber’s argument of the protestant work ethic. • Same critique as above. • Social norms are found to matter and may be hard to change. • Two types of protective arguments: Ø Social norms are institutions. Ø Social norms are created by institutions.

Why are some countries poor and other countries rich? • The ignorance hypothesis. Ø

Why are some countries poor and other countries rich? • The ignorance hypothesis. Ø Rulers do not know how to make poor countries rich. Ø ”By convincing rulers about what is good economics we can save the world”. Ø A&R (2012) argue against this view by saying that ”policymakers in poor countries get it wrong, not by mistake or ignorance but on purpose” (p. 68).

Why are some countries poor and other countries rich? • The institutional hypothesis. Poor

Why are some countries poor and other countries rich? • The institutional hypothesis. Poor countries are poor due to poor institutions. • Key question: Why not make the pie larger first and then have more to take from? • Because of commitment problems and since the distribution of resources affect political power.

And what are good institutions? • Those that make countries grow? • Property rights?

And what are good institutions? • Those that make countries grow? • Property rights? • Social norms? • Democracy?

The relationship between democracy and growth • Why would democracy affect growth? Ø Affects

The relationship between democracy and growth • Why would democracy affect growth? Ø Affects property rights. Ø Increases consumption and reduces investment. Ø Autocrats can defend themselves against (other? ) special interests. Ø Autocrats may have an easier time stealing.

Number of developing countries Success Autocracy 9 Democracy 1

Number of developing countries Success Autocracy 9 Democracy 1

The World Bank Growth Commission • “Growth at such a quick pace, over such

The World Bank Growth Commission • “Growth at such a quick pace, over such a long period, requires strong political leadership. ”

Autocracy Failure Neither Success 10 70 9 12 1 Democracy 0

Autocracy Failure Neither Success 10 70 9 12 1 Democracy 0

 • The argument suffers from a bias called: Ø ”Reversing conditional probabilities” Ø

• The argument suffers from a bias called: Ø ”Reversing conditional probabilities” Ø “Neglecting base rate bias” • Confuses the conditional probabilities P(A|B) and P(B|A). • The probability that you are an autocrat if you are a growth success is 90 percent. • However, the relevant probability is whether you are a growth success if you are an autocrat, which is only 10 percent.

Other biases (see Easterly 2011) • Why is the “benevolent autocrat” such a popular

Other biases (see Easterly 2011) • Why is the “benevolent autocrat” such a popular idea? • Availability heuristic. • Leadership attribution bias or fundamental attribution error. • The “Hot Hand” fallacy • The “Law of Small Numbers”

So, what is the relation? • In the short run it is obviously possible

So, what is the relation? • In the short run it is obviously possible to grow under autocracy. • But is it possible to become rich? • Acemoglu & Robinson (2012) basically argue that you need both property rights and broad based political power. • Seem to give primacy to the latter

Persistence and change • The best overview of the arguments is given in AJR

Persistence and change • The best overview of the arguments is given in AJR (2004), ”Institutions as the fundamental cause for long term growth”. • It is the manifesto of their research program. • The argument is divided into 6 parts.

 • Economic institutions (property rights, markets) shape incentives and therefore determine the growth

• Economic institutions (property rights, markets) shape incentives and therefore determine the growth potential but also the distribution of resources in the future. • The political power of groups determines what economic institutions will prevail. • The size of the pie is not maximized due to commitment problems. • Political institutions (democracy, budget rules) determine constraints and thereby de jure political power.

 • The distribution of resources affect the de facto political power (revolts, co-opt

• The distribution of resources affect the de facto political power (revolts, co-opt the military etc. ) • Political institutions are generally persistent.

Schematic framework Proximate cause State variables

Schematic framework Proximate cause State variables

2 sources of persistence • Political institutions are durable. • Relative richness tends to

2 sources of persistence • Political institutions are durable. • Relative richness tends to reproduce inequalities in both power and richness.

Change in this framework • There could be an exogenous shock in the distribution

Change in this framework • There could be an exogenous shock in the distribution of income (e. g. Atlantic trade, the Black death). • Due to different preconditions, these shocks play out differently. • For a revolution to be successful, it must change the political institutions (e. g. the glorious revolution).

Acemoglu et al. (2001) • Research question: Do institutions cause growth? Ø Interesting? Yes:

Acemoglu et al. (2001) • Research question: Do institutions cause growth? Ø Interesting? Yes: Central topic in development and they propose to have evidence for the mainstream view. ØOriginal? Yes, their causal channel has not been credibly tested before. Ø Feasible? Yes, by collecting innovative data and using IV.

The problem • Do institutions cause growth? Ø Richer countries may afford better institutions.

The problem • Do institutions cause growth? Ø Richer countries may afford better institutions. Ø Other factors may cause both growth and institutions. • Solution: Use an instrument for institutions.

Recap IV • To use the IV approach we need at least one additional

Recap IV • To use the IV approach we need at least one additional variable, referred to as an instrument. The instrument has to satisfy two conditions: • i) Relevance (easy to test) • ii) Validity (cannot be tested)

Settler mortality is argued to be key • They use data on the mortality

Settler mortality is argued to be key • They use data on the mortality rates of soldiers, bishops, and sailors stationed in the colonies. • The theory behind is that the settlers brought good institutions where they settled and extractive institutions in other areas.

Identification Strategy and argument (potential) settler mortality Settlements Early institutions Current economic performance

Identification Strategy and argument (potential) settler mortality Settlements Early institutions Current economic performance

Estimation

Estimation

GDP and institutions

GDP and institutions

Institutions and settler mortality

Institutions and settler mortality

Reduced form

Reduced form

Results • Both the OLS and IV results suggest that institutions are important for

Results • Both the OLS and IV results suggest that institutions are important for long run growth. • The intermediate steps are also found to be consistent with their theory. • The first stage shows that the instrument is relevant.

Validity • “conditional on the controls included in the regression, the mortality rates of

Validity • “conditional on the controls included in the regression, the mortality rates of European settlers more than 100 years ago have no effect on GDP per capita today, other than their effect through institutional development. ” • Cannot be confirmed, only rejected.

Validity • What about current disease environment? • “We believe that this is unlikely

Validity • What about current disease environment? • “We believe that this is unlikely to be the case and that our exclusion restriction is plausible. ” • Arguments: Ø Deaths mainly due to malaria and yellow fever: Indigenous adults are immune. Ø Robust to controlling for current disease environment and infant mortality. ØSimilar results with yellow fever instrument (which is mostly eradicated today).

Validity • What about other channels? • They do overidentification tests. • ”However, such

Validity • What about other channels? • They do overidentification tests. • ”However, such tests may not lead to a rejection if all instruments are invalid, but still highly correlated with each other. Therefore, the results have to be interpreted with caution. ” • We will return to this in Glaeser et al. (2004)

Glaeser et al. (2004) • Research question: Do institutions (really) cause growth? Ø Interesting?

Glaeser et al. (2004) • Research question: Do institutions (really) cause growth? Ø Interesting? Yes: Central topic in development and they propose to have evidence against the mainstream view. Also important critical discussion on measurement issues. ØOriginal? Yes, their causal channel has not been credibly tested before. Ø Feasible? Yes, basically a replication study.

A familiar problem • Do institutions cause growth, or is it growth (in income

A familiar problem • Do institutions cause growth, or is it growth (in income or human capital) that causes institutional improvement? • They start with the example of North and South Korea.

North versus South Korea In 1980 SK had double per capita income as compared

North versus South Korea In 1980 SK had double per capita income as compared to NK

Institutions • ”A set of rules, compliance procedures, and moral and ethical behavioral norms

Institutions • ”A set of rules, compliance procedures, and moral and ethical behavioral norms designed to constrain the behavior of individuals. . . ”(North 1981). • The constraints should also be permanent or at least durable.

Three common measures of institutions • Institutional quality from ICRG. Subjective assessments of risks

Three common measures of institutions • Institutional quality from ICRG. Subjective assessments of risks faced by investors. • Government effectiveness from Kaufmann et al. (2003). Also subjective assessments. • Polity IV from Jaggers and Marshall (2000). Measure constraints on the executive and democracy.

Forceful critique • ”. . . the commonly used measures cannot be used to

Forceful critique • ”. . . the commonly used measures cannot be used to establish causality. ” • They reflect outcomes and choices. • Do not measure constraints on government. • They are highly volatile. • Only barely correlated with objective measures of electoral rules.

Alternative hypothesis • Test the hypothesis that education is driving economic development. • First

Alternative hypothesis • Test the hypothesis that education is driving economic development. • First of all, initial level of education is an equally strong predictor as the commonly used measures of institutions. • And, the objective measures of institutions are not significant.

Objective measures

Objective measures

Reverse causality of institutions • Assessments may improve as the country gets richer, so

Reverse causality of institutions • Assessments may improve as the country gets richer, so that causality runs the other way. • It is more difficult to argue that economic growth in e. g. the 70 s affect education in the 60 s. • So let us turn to the IV results of Acemoglu et al. (2001).

Critique against the settler mortality instrument • Even if one agrees that mortality risk

Critique against the settler mortality instrument • Even if one agrees that mortality risk shaped settlement decisions, how do we know that it is institutions that matter? • They also brought with themselves and hence their human capital. • If settlement affects growth via human capital, the instrument is not valid.

Critique against the settler mortality instrument • Settler mortality is uncorrelated with objective measures

Critique against the settler mortality instrument • Settler mortality is uncorrelated with objective measures of institutions. • Settler mortality is correlated with the modern disease environment. • They present their own IV regressions with settler mortality as instrument either for schooling or institutions.

Results • Years of schooling, but not executive constraints, are statistically significant. • May

Results • Years of schooling, but not executive constraints, are statistically significant. • May still be other things the Europeans brought with them like, ”Guns, germs, and steel”. • The point is, why should we think that it is institutions that drives it?

The debate is ongoing • Very good discussion over at the IIES: • http:

The debate is ongoing • Very good discussion over at the IIES: • http: //urplay. se/Produkter/172344 -URSamtiden-Nobelsymposium-Tillvaxt-ochutveckling-Institutioner • Here the authors actually meet and Acemoglu gives an answer to the critique.

Bruhn and Gallego (2012) • “GOOD, BAD, AND UGLY COLONIAL ACTIVITIES: DO THEY MATTER

Bruhn and Gallego (2012) • “GOOD, BAD, AND UGLY COLONIAL ACTIVITIES: DO THEY MATTER FOR ECONOMIC DEVELOPMENT? ” • Conclusion: Bad colonial activities (labor exploitation) led to lower economic development today. “Our results also suggest that differences in political representation (but not in human capital) could be the intermediating factor between colonial activities and current development. ”

Heldring and Robinson (2012) • ”We argue that in the light of plausible counterfactuals,

Heldring and Robinson (2012) • ”We argue that in the light of plausible counterfactuals, colonialism probably had a uniformly negative effect on development in Africa. ”