Journal of Economic Literature 2015, 53(4), 898–944 http://dx.doi.org/10.1257/jel.53.4.898 Culture and Institutions† Alberto Alesina and Paola Giuliano* A growing body of empirical work measuring different types of cultural traits has shown that culture matters for a variety of economic outcomes. This paper focuses on one specific aspect of the relevance of culture: its relationship to institutions. We review work with a theoretical, empirical, and historical bent to assess the presence of a two-way causal effect between culture and institutions. ( JEL D02, D72, I32, J12, Z13) 1. Introduction we also sum up how various authors have defined them differently. ecent research demonstrates that cul- Culture and institutions are endogenous Rtural variables determine many eco- variables determined, possibly, by geography, nomic choices—they even affect the speed technology, epidemics, wars, and other his- of development and the wealth of nations.1 torical shocks. Can any causal link between Researchers are now striving to better the two be established? How do culture and understand the mechanisms. institutions interact? In this paper, we investigate what we know One notable study—by Putnam (1993), about one specific mechanism: the relation- on social capital in Italy—illustrates how ship between culture and institutions. Both complex these issues are. Putnam and his terms are often vague in the literature; we colleagues took advantage of a natural exper- devote space to defining them properly, and iment involving an institutional reform: in the early 1970s, Italy’s central government established fifteen new regional govern- ments.2 Ideally, they would function identi- * Alesina: Harvard University and IGIER Bocconi. Giuliano: University of California, Los Angeles, Anderson cally throughout the country, but in practice School of Management. We thank Benjamin Friedman and they didn’t. The discrepancy was most pro- Andrei Shleifer for useful conversations and Janet Currie, nounced between the center-north and the Steven Durlauf, and six anonymous referees for excellent comments. south. Putnam and his colleagues hypothe- † Go to http://dx.doi.org/10.1257/jel.53.4.898 to visit the sized that the variance was due to regional article page and view author disclosure statement(s). differences in levels of cooperation, partic- 1 Several economics papers have investigated what cul- tural traits are relevant for development, their persistence, ipation, social interaction, and trust—four and their historical origins. Several surveys have analyzed some of these aspects (see Guiso, Sapienza, and Zingales 2006 and Fernandez 2008, 2011). For an informal treat- 2 The reform, which implemented an article originally ment of the question of how cultural values affect develop- approved in the 1945 constitution, can be reasonably con- ment, see Landes (1998). strued as independent of regional development. 898 Alesina and Giuliano: Culture and Institutions 899 key “social capital” traits. They argued that document below, they try as well as they can these regional differences—dating back at to isolate the effect of culture from institu- least as far back as the twelfth century—are tions—probably because the importance of a function of whether the given region had institutions is fairly well established.3 Since experienced the institution of free cities. cultural economics is in its infancy, those Free cities developed a form of early par- who write about institutions don’t seem to ticipatory democracy, generating a feeling worry much about whether institutions are of belonging to a polity, whose function- well identified and isolated from cultural ing could guarantee both protection from influences, which may be problematic. Some aggression and the provision of public goods. may argue that culture is a vague variable As a result, citizens of free cities developed a and difficult to measure. One of our ancil- deep sense of civic and cooperative behavior, lary goals here is to try to clarify these defi- a cultural trait they transmitted from genera- nitional issues. tion to generation. The rest of our paper is structured as fol- Guiso, Sapienza, and Zingales (2013) lows. In section 2, we define what culture formally tested this hypothesis, finding means in the economic literature, and how considerable support for it. A contempo- it is measured. Many contributions to the rary Italian city’s social capital, a “cultural” literature since the last two surveys discuss variable determining the success or fail- the relevance of culture on economic out- ure of its institutions, correlates with its comes. Thus, we provide a map of the main historical experience as a free city in the cultural traits used in economics and their Middle Ages. Thus, an institutional vari- correlations. We also provide definitions able, the free-city arrangement, influenced a and measurements of formal institutions. long-lasting cultural change that still affects In section 3, we scrutinize the relationship Italy’s local governments. If cultural values between culture and institutions, first by were not so persistent, being a free city in reviewing existing empirical and theoretical the twelfth century would have nothing do literature that shows how culture can affect with today’s institutions. At the same time, formal institutions, and then by reviewing this long-lasting cultural trait was sparked recent studies that show how formal institu- by early forms of local self-determination, an tions affect culture. Then, we document the institutional feature. interplay between culture and formal insti- The experience of a free city in the Middle tutions and review the literature on how they Ages is clearly not an exogenous variable. jointly determine economic development. For example, even within central and north- ern cities, there is variation regarding which 2. Definitions and Measurement cities could more easily become free, due to of Culture and Institutions geographic features that made them more or less capable of defending themselves against 2.1 Definitions of Culture the emperor. Like geography, many other factors could have determined the relative Defining culture is an arduous task. We efficiency of local governments in Italy. Yet start by providing a definition, distinguish- the complex interaction between culture and ing between empirical and theoretical institutions is interesting, regardless of the “ultimate” causes. 3 Various controversies remain regarding how, where, Those who study culture are well aware and in what sense institutions matter. See Glaeser et al. of the importance of institutions and, as we (2004). 900 Journal of Economic Literature, Vol. LIII (December 2015) definitions of culture. The reason for the dis- culture, an important component of which is tinction is that the mapping between empir- cultural beliefs. Cultural beliefs are the ideas ical and theoretical concepts is often not and thoughts common to several people that straightforward. govern interaction—between these peo- On the empirical side, most papers (if not ple and among them, their gods, and other all) follow the definition adopted by Guiso, groups—and differ from knowledge in that Sapienza, and Zingales (2006), where cul- they are not empirically discovered or ana- ture is defined as “those customary beliefs lytically proved. In general, cultural beliefs and values that ethnic, religious, and social become identical and commonly maintained, groups transmit fairly unchanged from gen- and communicated.” Greif asserts there is eration to generation.” Empirical papers, a subset of “rational cultural beliefs, which therefore, combine values and beliefs in the capture people’s expectations with respect same definition. to actions that others will take in various On the theoretical side, values and beliefs contingencies. Past cultural beliefs that are often treated differently. Several authors sustain Nash equilibria provide focal points have developed models in which culture in repeated social interactions or when there means beliefs about the consequences of are multiple equilibria.”4 one’s actions, but where these beliefs can Still others view culture as a more prim- be manipulated by earlier generations or itive phenomenon embodied in values and by experimentation. For example, Guiso, preferences (see, for example, Akerlof and Sapienza, and Zingales (2008a) show how Kranton 2000). This definition, also used individual beliefs are initially acquired in psychology (Pinker 1997; Kaplow and through cultural transmission and then slowly Shavell 2007) emphasizes the role of emo- updated through experience, from one gen- tions in motivating human behavior. eration to the next. They use an overlapping The two interpretations are not mutu- generation model in which children absorb ally exclusive. Benabou (2008) shows that their trust priors from their parents and then, values and beliefs interact systematically. after gaining real-world experience, transmit He incorporates “mental constructs” into their updated beliefs to their own children. a political-economy model and shows that In this setting, multiple equilibria are pos- these mental constructs interact with insti- sible. In the no-trust-no-trade equilibrium, tutions to generate different beliefs, which beliefs of mistrust are transmitted from par- could persist over time. ents to children, who eventually shun trade Empirical investigation of the relevance of and therefore do not learn about the trust- culture on economic outcomes is fairly new