Correlation And/Or Causality? Recent Empirical Approaches ØRecent years witnessed a soaring interest in empirically assessing the effect of institutions on long-run economic development/growth ØResearch provides an important step forward since the seminal work by North and Thomas ØEmpirical challenges partially resolved, but new challenges emerged ØWe are going to critically examine empirical approaches used in economics to estimate the role/impact/effect of institutions on various econmic outcomes Empirics of Institutions - why? ØInstitutions are rules of the game in a society or, more formally, are the humanly devised constraints that shape human interactions (North 1990, page 3) ØMain points: ØHumanly devised ØSet constraints ØShape incentives ØThe economic concept of institution is defined as constraints placed by law and social norms on human behavior ØThese constraints help to reduce transaction costs Ø Institutions – what, why? ØIt is a broad cluster which includes many sublevels, e.g. property rights, contract enforcement ØFormal institutions: codified rules by law, e.g. in the constitutions ØInformal institutions: not legally codified, based on social norms (or conventions), accepted and expected standards of behaviour ØInformal institutions are based on culture, the set of beliefs and values passed from generation to generation Institutions – what, why? ØWhy do we (should we) care? ØVast differences in economic prosperity among countries ØE.g. income per capita in sub-Saharan countries is on average 1/20th of per capita income in the United States ØWhy??? ØStandard answer offered by economists: ØDifferences in physical capital ØDifferences in human capital ØDifferences in technology Institutions – what, why? ØNorth and Thomas (1973) argue that we need to distinguish between proximate causes versus fundamental causes of prosperity ØProximate causes are capital, technology ØFundamental causes are can be ØInstitutions ØGeography ØCulture fundamental causes proximate causes economic development Ø Institutions – what, why? ØThe issue with institutions is that they are endogenous and develop in tandem with other determinants of economic development ØInstitutions can be different because of ØGeography ØCulture ØOther factors ØAn old story by Montesquieu: ØGeography determines human attitudes ØHuman attitudes determine both economic performance and political system ØSo institutions can be also determined by human attitudes Ø Institutions – an empirical problem ØWhy is it a problem? ØCausality vs Correlation ØTo estimate a causal relationship between institutions and economic development Yi=α + β*xi + γ*institutionsi + εi ØAssuming that cov(xi, εi)=0, argument made earlier implies that cov(institutionsi, εi)≠0 Ø=> we can’t interpret the estimated coefficient of γ as a causal relationship between institutions and economic development, but only as a correlation Institutions – an empirical problem ØVarious approach which attempt to isolate plausibly exogenous sources of variation of institutions ØTwo main strategies: 1.Finding institutions which exogenously vary across space because of ØGeography ØHistorical events 2.Instrumental variable estimation techniques ØFinding a variable zi which is highly correlated with a variable proxying institutions and not correlated with εi a.cov(zi, εi)=0; can’t be tested (often called exclusion restrictions) b.cov(zi, institutionsi) ≠ 0; can be tested 3.Combination of point 1 and point 2 a. Solutions ØAttempt to examine and establish a causal relationship between institutions and per capita income ØTo do that, they need a source of exogenous variation in institutions ØMain arguments rest on: 1.Different types of colonization policies created different set of institutions 2.The colonization strategy was influenced by the feasibility of settlements 3.The colonial institutions persisted after independence AJR (2001): Colonial Origins of Comparative Development ØTwo extremes of colonization strategies: ØExtractive institutions to transfer resources from the colony to the mother country which led to the creation of extractive institutions e.g. coerced labor, slavery, monopolies, legal discrimination ØInclusive institutions which replicated European institutions more conducive for economic growth ØThe colonization strategy was influenced by mortality rates expected by the first European settlers (feasibility of settlement) AJR (2001) Potential mortality of settlers Colonization strategy Past colonial institutions Current institutions Current performance AJR (2001) AJR (2001) ØAssumption of exogeneity of settlers’ mortality (exclusion restriction): ØConditional on other controls, the settlers’ mortality rates have no effect on GDPpc today other than through institutional development ØThe main concern is that mortality rates of settlers are correlated with current disease environment which can have a direct effect on economic performance ØIf it was a case, settlers mortality would be assigning the effect of disease on income to institutions ØAJR argue it is not the case as diseases were fatal for the Europeans but not so much for the indigenous population AJR (2001) ØPersistence of colonial institutions ØSetting up and enforce institutions is costly – if the costs were sunk by the colonial powers, then it may not be beneficial for the elites after independence to change them ØGains from extractive institutions may depend on the size of the ruling elite; if the elite is small (often the case), it has no incentive to switch to ‘inclusive institutions’ ØIf irreversible investments which are complementary to a particular set of institutions were made, those who made them are more likely to make those institutions to persist AJR (2001) ØVariables capturing institutions: ØIndex of protection against expropriation • • • • AJR (2001) cov(εi, νi)=0 AJR (2001) AJR (2001) ØDell, M (2014): Path Dependence in Development: Evidence from the Mexican Revolution, mimeo Harvard Øimpact of insurgency on income, labor force, public employees, education ØInstrumental variable approach in which insurgency is instrumented with severity of drought • Other studies cov(εms, μms)=0 ØIyer, Lakshmi (2010): Direct versus Indirect Colonial Rule in India: Long-Term Consequences, The review of Economics and Statistics 92(4), 693-713 ØEffect of direct vs indirect rule of the British empire on various economic outcomes in India ØProblem of endogeneity of direct British rule ØDoctrine of annexation – annex an Indian territory is a ruler died without a natural heir Ø Ø Other studies cov(εi, ui)=0 ØAssumption of ‘exclusion restriction’ can’t be tested => there is a scope for ‘residual’ endogeneity ØGeography and history can provide a much ‘cleaner’ way of estimating a causal impact of institutions on various economic outcomes ØTwo papers will be discussed: ØBasten, C., and Frank Betz (2013): Beyond Work Ethic: Religion, Individual and Political Preferences, American Economic Journal: Economic Policy 5, 67-91 ØDell, M. (2010):the Persistent Effects of Peru’s Mining Mita, Econometrica 78(6), 1863-1903 Geography and history as a source of exogenous variation ØExamine the effect of Reformed Protestantism (relative to Catholicism) on preferences for leisure, redistribution, and intervention in the economy ØUse data from Switzerland which allows for within-country correlation and considerable geographical and institutional variation => concentrate on institutionally homogenous part of South-Western Switzerland (Vaud and Fribourg) ØArgue that Reformation led to exogenous variation in religion BB (2013) ØExogeneity of religion is historically argued: ØHomogenous until the early 16th century ØThen split into two parts – one belonged to Berne, the other to Fribourg ØBoth imposed different religion: Berne Protestant, Fribourg Catholic ØHomogeneity of both regions before the adoption of Protestantism is crucial for a valid causal analysis ØProblem with the main explanatory variable – share of Protestants – changes discontinuously at the historical border between Catholic and Protestant parts ØAt the same time, it is exogenous change, offering us a possible instrument – distance to the historical border BB (2013)