Rethinking Economic Development PDF

1 Rethinking economic development Nathan Nunn Department of Economics, Harvard University Abstract. I provide a summary, reﬂection and assessment of the current state of economic development in both the policy and academic worlds. In terms of development policy, currently, the primary focus is on policy interventions, namely, foreign aid, aimed at ﬁxing the “deﬁciencies” of developing countries. Academic research also has a similar focus, except with an emphasis in rigorous evaluation of interventions to estimate causal eﬀects. A standard set of versatile quantitative tools is used, e.g., experimental and quasi- experimental methods, which can be easily applied in a range of settings to estimate the causal eﬀects of policies, which are typically presumed to be similar across contexts. In this article, I take a step back and ask whether the current practices are the best that we can do. Are foreign aid and policy interventions the best options we have for poverty alleviation? What else can be done? Is our current research strategy, characterized by rigorous but a lack of context-speciﬁc analysis, the best method of analysis? Is there a role for other research methods, for a deeper understanding of the local context and for more collaboration with local scholars? Résumé. Repenser le développement économiqueé. Dans cet article, je propose un aperçu, une réﬂexion et une évaluation de l’état actuel du développement économique à la fois dans les sphères politiques et universitaires. En matière de développement économique, l’attention est principalement portée sur l’intervention politique, c’est-à-dire l’aide extérieure visant à corriger les défaillances des pays en développement. Les recherches universitaires y portent une attention similaire, mais en mettant davantage l’accent sur l’évaluation rigoureuse de ces mêmes interventions pour en mesurer les eﬀets de causalité. À cette ﬁn, un ensemble d’outils quantitatifs et modulables standard sont utilisés, notamment par l’entremise de méthodes expérimentales et quasi expérimen- tales pouvant facilement s’appliquer à une vaste gamme de situations, et ce aux ﬁns d’évaluations des liens de causalité engendrés par ces politiques (liens habituellement considérés comme similaires d’un contexte à l’autre). Dans cet article, je prends un peu de recul pour m’interroger sur la pertinence des pratiques actuelles. Les interventions politiques et l’aide extérieure oﬀrent-elles les meilleures options pour réduire la pauvreté? Que pouvons-nous faire d’autre? Notre stratégie de recherche actuelle, caractérisée par des analyses rigoureuses mais peu adaptées à des contextes spéciﬁques, est-elle la meilleure? D’autres méthodes de recherche peuvent-elles permettre de mieux comprendre les contextes régionaux et favoriser une meilleure collaboration avec les intellectuels locaux? Corresponding author: Nathan Nunn, [email protected] The article is based on the Innis Lecture delivered at the 53rd Annual Conference of the Canadian Economics Association, which was held in Banﬀ Alberta, May 31–June 2, 2019. For helpful comments and discussions, I thank an anonymous referee as well as Grieve Chelwa, Sahana Ghosh, Joseph Henrich, Jacob Moscona, Suresh Naidu, James Robinson and Jessica Trisko Darden. I also thank Lydia Cox and Aditi Chitkara for excellent research assistance. Canadian Journal of Economics / Revue canadienne d’économique 20XX 000(0) XXX 2019. Printed in Canada / XXX 2019. Imprimé au Canada ISSN: 0008–4085 / 19 / pp. 1–25 / © Canadian Economics Association 2 N. Nunn JEL classiﬁcation: O19, O21, O25 1. Introduction U U nderstanding how to eradicate global poverty and improve the economic well-being of the poorest communities in the world is one of the most important issues facing academics and policy-makers today. In the policy world, the ﬁght against global poverty has focused on foreign aid and policy interventions aimed at helping developing countries obtain sustained economic growth. In academia, development research also emphasizes policy interventions, but with the focus on the rigorous evaluation of their eﬀects. This is most commonly achieved using randomized control trials (RCTs), where the policy intervention is provided to a randomly chosen treatment group leaving a comparable and randomly chosen control group. The random- ization provides an easy way to obtain credible estimates of the causal eﬀects of the intervention. In the past decades, there has been an important revolution. It is now recognized that credible evaluations of policy interventions are crucial and that the gold standard method to achieve this is RCTs. Given how diﬃcult it is to see the bigger picture when one is in the thick of things, it is never a bad idea to step back and evaluate the current state of the thinking and practice of economic development in both the policy world and the academic world. Part of this process involves asking ourselves whether we could be doing anything better. Are there ways in which our current thinking and practice could be improved? It would be odd if there were no areas in which we could be doing better. The goal of this article is to highlight and discuss these and to oﬀer thoughts about alternatives. 2. Rethinking policy 2.1. Foreign aid One of the key tools used to alleviate poverty is foreign aid, which is the trans- fer of money, commodities and services from a foreign country or international organization for the beneﬁt of the recipient country’s population. It takes the form of grants or concessional loans and is typically classiﬁed into economic aid, military aid and humanitarian (emergency) aid. The most common form of aid is oﬃcial development assistance (ODA), which is primarily comprised of bilateral grants. When using the term here, I am also including transfers across borders by foundations, religious organizations and NGOs, but not remittances between friends and family. On the ground, foreign aid is used to fund a range of projects that take a wide variety of forms. There is evidence that many of these projects have siz- able beneﬁts. Numerous studies convincingly identify the beneﬁts of a range of interventions that were made possible through foreign aid. For example, we know that the deworming program, which was organized and implemented Rethinking economic development 3 by the Dutch NGO Internationaal Christelijk Steunfonds Africa (ICS) in Busia Kenya, led to school attendance and higher adult wages Miguel and Kremer (2004), Baird et al. (2016). We also know that in Kenya, large-scale unconditional cash transfers given by the US-based NGO GiveDirectly (GD) led to an increase in household assets, monthly revenues and psychological well-being (Haushofer and Shapiro 2016, 2018). At the same time, there are many who voice concerns that foreign aid may have unintended consequences. Because these eﬀects are unintended, they are generally not unexpected or foreseen and, thus, not evaluated. Thus, we may be systematically underestimating the adverse eﬀects of foreign aid. Given this, whether aid can have adverse consequences is an important question. A priori, there are many reasons to think that this is a possibility. First, there is evidence that aid is often determined by the strategic or economic needs of the donor countries (e.g., Alesina and Dollar 2000, Kuziemko and Werker 2006). One indicator of this is the fact that historically the majority of foreign aid has been “tied,” which means that the concessional loans or grants that are given come with the requirement that the money has to be used to purchase speciﬁc products, namely, those produced by the donor country. The United States has and continues to be one of the countries with the highest proportion of tied aid. While the tying of aid is beneﬁcial for the donor country, it is eﬀectively a form of export promotion that poses signiﬁcant costs for the recipient country. It is estimated that the tying of aid raises the costs of goods and services (i.e., decreases value for money) by 15% to 30% on average and by 40% for food aid (Clay et al. 2009). We also know that a large proportion of aid can go missing. One of the earliest studies to track and document missing foreign aid funds was Reinikka and Svensson (2004), who studied World Bank funded capitation grants in Uganda that were supposed to pay for all education-related expenses other than teacher’s wages and buildings. The authors developed a way of measuring the amount of the funds that actually reached the school. They found that the median school in their sample received 0% of its funds (i.e., 100% was stolen). Across all schools, on average, 87% of the funds went missing. Using the same protocol, they also ﬁnd similarly high rates of fund capture elsewhere: 49% in Ghana, 57% in Tanzania and 76% in Zambia. Olken (2007) applied the theft- measurement strategy used by Reinikka and Svensson (2004) to road main- tenance aid projects and found that approximately 30% of funds went missing. Having such resources “up for grabs” potentially aﬀects the incentives of the most entrepreneurial and most talented in a country, increasing the likelihood that they engage in zero-sum rent-seeking activities rather than productive activities that are beneﬁcial for the country as a whole (Bhagwati 1982). Although empirically estimating convincing causal eﬀects of foreign aid on corruption remains elusive due to measurement and identiﬁcation chal- lenges, given the ﬁndings of Reinikka and Svensson (2004) and Olken (2007), it is almost tautological that aid increases corruption. Given that aid increases the amount of funds available, a fraction of which are stolen through corrupt 4 N. Nunn activities, then this must increase corruption if it is deﬁned to be those activ- ities that were used to capture the funds. The only way in which aid would not increase corruption in this scenario is if 100% of the eﬀort to capture aid funds would have gone to other corrupt activities, which seems unlikely. Of course, it is possible that although aid may increase corruption today, because it increases economic growth it will reduce corruption in the future. However, the evidence for growth eﬀects of foreign aid remains elusive (Werker et al. 2009). There are many examples of studies ﬁnding adverse consequences of foreign aid. For example, Haushofer et al. (2015) estimate the eﬀects of unconditional cash transfers (UCT) on the neighbours of recipients. They show that one’s happiness is reduced when one’s neighbour receives an UCT. Interestingly, this is true whether or not the individual also receives a cash transfer. Werker et al. (2009) exploit variation in oil prices that drives aid contributions of oil- rich OPEC countries in the Middle East to less-developed Muslim countries around the world. They ﬁnd that foreign aid does not aﬀect investment or GDP growth. However, it does lead to a signiﬁcant increase in household and government consumption, which primarily takes the form of increased imports of non-capital products. Thus, foreign aid does not fuel growth-promoting investments (or growth itself) but instead crowds out domestic savings and increases consumption of foreign products. One of the most important potential adverse consequences of foreign aid is its inﬂuence on conﬂict. There are many accounts of foreign aid fuelling conﬂict. One such case is believed to have occurred during the Nigeria–Biafra civil conﬂict of the late 1960s (Barnett 2011, pp. 133–47). After a year of conﬂict, Biafra—the region of Nigeria seeking independence—was surrounded and being closed in on by the Nigerian army. The siege led to starvation. After the media reported on this, there was a global outcry for humanitarian relief. Figure 1 provides an example of this, showing the cover of Life magazine from July 12, 1968. Although opposed by the Nigerian government, humanitarian aid was brought into Biafra. According to some accounts, the Biafran leader, Odumegwu Ojukwu, initially allowed aid to enter the rebel-controlled region of Biafra only if it was shipped on his planes. He charged for this service and also ﬁlled them with arms and other military equipment. Even when organizations were allowed to use their own planes, they were still charged landing fees and import taxes. The shipments of humanitarian aid allowed Ojukwu to circumvent the siege and to feed and arm his troops. Many argue that the shipment of humanitarian aid resulted in the Biafran civil conﬂict lasting years longer than it would have otherwise (Polman 2010, pp. 115–19). Another commonly cited example is the experience of the Eastern DRC following the Rwandan Genocide in 1994. Hutu extremist leaders, who had ﬂed Rwanda, took refuge in humanitarian camps in the DRC, where they taxed Hutu civilians in the camps and transferred the appropriated aid to their militia. The aid and physical protection provided by refugee camps allowed the Hutu extremists to regroup and rebuild their army. The Hutu militia was then able to carry out raids into Rwanda. The Rwandan government Rethinking economic development 5 FIGURE 1 Cover of Life magazine, July 12, 1968 responded by arming the Banyamulenge (ethnic Tutsis living in the Eastern DRC), which led to the Banyamulenge Rebellion in 1996, which was then followed by the ﬁrst and second Congo wars (Terry 2002, ch. 5; Lischer 2005, ch. 4). The region of the DRC has been in a state of persistent conﬂict and insecurity ever since. Numerous studies have formally tested for relationships between foreign aid and conﬂict, using a range of identiﬁcation strategies to obtain credible causal estimates. Many studies ﬁnd that foreign aid can increase conﬂict. Nunn and Qian (2014) ﬁnd this to be the case for US food aid. Their analysis uses an IV strategy where US wheat production shocks, combined with a country’s tendency to receive wheat aid from the US to obtain exogenous variation in US food aid supply. Crost et al. (2014) use an RD strategy that exploits an eligibility cut-oﬀ for a World Bank funded development program in the Philippines to estimate the eﬀects of the program on conﬂict. They ﬁnd that eligibility to participate in the program is associated with more conﬂict, which appears to be due to an increase in insurgent attacks against government forces in an attempt to disrupt the program. Dube and Naidu (2015) estimate the eﬀects of military aid in Colombia using a diﬀerences-in-diﬀerences identiﬁcation strategy. They ﬁnd that US military aid leads to an increase in conﬂict and violence arising due to an increase in attacks by paramilitaries. 6 N. Nunn Others have argued that foreign aid is often channeled towards strength- ening the government’s military, particularly in autocratic regimes where stability requires oppression and the use of force (Kono and Montinola 2009). While it is unsurprising that military aid might have such an eﬀect, this has actually been found for economic aid. Because of the fungibility of economic aid, it is indirectly channeled towards military and violent ends. This occurs through the sale of humanitarian aid (e.g., food aid) for cash and by freeing up more of the government budget for military purposes. The recent analysis by Trisko Darden (2020) shows both through case studies and empirical analysis that this eﬀect is commonly observed. The author shows that the year following an increase in US foreign aid into a country, there is an increase in killings, repression and torture by the state. This echoes an earlier ﬁnding by Ahmed (2016): increased US foreign aid is associated with greater repression and human rights abuses in the recipient country. While there is evidence that foreign aid can increase conﬂict, it is not the case that it always leads to conﬂict. For example, Nunn and Qian (2014) show that among the countries in their sample without a recent history of past conﬂict, food aid does not increase conﬂict. In a follow-up study that studies a conditional cash transfer program also in the Philippines, Crost et al. (2016) ﬁnd that this aid package actually decreased conﬂict. Trisko Darden (2020) ﬁnds that the eﬀect of US aid on state killings and repression of its citizens is weaker following the end of the Cold War. Given the evidence that foreign aid can sometimes cause conﬂicts, but at other times have no eﬀects or even reduce conﬂict, the natural next question is how we implement foreign aid projects in a manner that minimizes its harm, thus maximizing overall beneﬁt. A recent study by Moscona (2019) takes an important step in this direction by estimating the heterogeneous eﬀects of the universe of World Bank lending projects from 1995–2014. He ﬁnds that the eﬀect of aid on conﬂict is greater the less successfully the project was implemented, as measured by project quality scores given by the World Bank Independent Evaluation Group. Put diﬀerently, he ﬁnds that better run projects generate less conﬂict. According to the estimates, having the most poorly implemented aid projects results in more conﬂict than having no aid projects and having the best-implemented aid projects results in less conﬂict than having no aid project. These ﬁndings nicely sum up the message that has emerged from anecdotal examples and from the case study literature (e.g., Anderson 1999). Foreign aid if implemented successfully, can foster development and reduce conﬂict. However, if poorly implemented, it can exacerbate the situation and lead to more conﬂict. 2.2. Other policies? The fact that foreign aid is no panacea for underdevelopment raises the natural question of whether there are other feasible policy options. A commonly used Rethinking economic development 7 tool by countries has been trade and industrial policies. The underlying logic of the policy is motivated by the fact that there is a remarkably strong relationship between a country’s level of economic development and what it produces (Hausman et al. 2007). Without exception, countries producing sophisticated high-end manufactures, like automobiles and electronics, are relatively wealthy. Given this relationship, many countries have taken steps to promote their specialization in these “winner industries.” These include subsidies, low-interest loans and tariﬀ protection for producers in these sec- tors. Evidence is now accumulating that successful implementation of these policies can promote size and even the productivity of these industries, help- ing to jump-start industrialization and economic development. We now have evidence of the long-run success of such policies from 19th-century France (Juhasz 2018), late 19th- and early 20th-century Canada (Harris et al. 2015), late 20th-century South Korea (Lane 2017), post-WWII Italy (Giorcelli 2019) and post-WWII Finland (Mitrunen 2019). There is an important shortcoming of industrial policy, which becomes clear once one considers the fallacy of composition. Although each country can gain through this strategy, this is not true for developing countries as a whole. That is, every country can’t be a South Korean success story. This is because much of the gains from industrial policy come at the expense of other countries. Much of the beneﬁts of this strategy are zero sum. This raises the question of whether this is a successful strategy for economic development. The answer appears to be a clear no, at least not for most countries. However, the fact that these beneﬁts are zero sum provides insight into what can be done to help developing countries. To see this, consider antidumping duties. Due to multilateral tariﬀ reduc- tions by countries who have joined the GATT/WTO, tariﬀs have declined signiﬁcantly around the world. However, there are cases when countries can obtain exceptions and apply sizable barriers against foreign producers. An- tidumping duties are one example of these. In theory, they are supposed to be applied only in cases where a foreign producer is engaging in “dumping,” namely, pricing goods below cost in an eﬀort to drive all competitors from the market, after which they will then engage in monopoly pricing. In reality, anti- dumping duties have very little to do with anti-competitiveness and everything to do with traditional trade protection to protect domestic producers. They have proven to be an eﬀective tool. Antidumping duties are large, on average 10–20 times higher (and as much as 100 times higher) than MFN tariﬀs. On average, the duties cause the value of imports to fall by 30% to 50%. Strikingly, this decline in imports is found even if a case is ﬁled (i.e., initiated) but the duties themselves are never levied (Prusa 2001). Thus, the threat of the duties itself can have sizable eﬀects. Given these great features of anti-dumping duties, it is not surprising that at the same time that tariﬀs have declined globally there has been a rise in anti-dumping duties. This increase, which is shown in ﬁgure 2, has been quite rapid, especially during the 1980s and 1990s. 8 N. Nunn Total Initiations 350 300 250 Total Measures Number of Cases 200 150 100 50 0 1978 1982 1986 1990 1994 1998 2002 2006 2010 Year FIGURE 2 Total number of new anti-dumping initiations and measures each year from 1978–2013 SOURCE: Global Antidumping Database Since the initiation of antidumping duties requires signiﬁcant legal cap- acity, these are typically initiated by wealthier countries and are often against less-developed countries. This is shown by ﬁgure 3, which plots net initiations (i.e., initiations by a country minus initiations of other countries against the country from 1978–2013) and the country’s average income (measured in 1993). One observes a clear positive relationship. Those initiating duties against other countries tend to be wealthier while those that have duties placed against their products tend to be poorer countries. While the aggregate eﬀects of these duties for developing countries are not fully understood, we do have some evidence from one of the hundreds of duties that have been ﬁled. This is for an anti-dumping duty that was placed against Vietnamese catﬁsh farmers by Mississippi catﬁsh farmers in 2003. The eﬀects on Vietnamese households has been studied by Brambilla et al. (2012). They ﬁnd that for Vietnamese households that had specialized in catﬁsh farming, average annual real per capita income decreased by 40%. This is striking given that, as shown by ﬁgure 2, there are hundreds of new antidumping duties imposed each year. This is an enormous eﬀect and much larger than the size of any eﬀect on incomes found for any form of foreign aid or any policy intervention. If Rethinking economic development 9 500 USA 400 European Union Initiations by − Initiations against 300 Australia Canada 200 India Argentina South Africa Brazil 100 Mexico PakistanPeru Colombia Israel New Zealand Costa Ecuador Rica Russia Chile Venezuela Malaysia Japan 0 Indonesia Thailand South Korea −100 China −200 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 GDP per capita (PPP) FIGURE 3 Relationship between net anti-dumping initiations and average per capita GDP NOTE: The correlation coeﬃcient between the two variables is 0.55, which has a p-value of 0.004. SOURCE: Global Antidumping Database we did ﬁnd an intervention that increased real per capita incomes by 40% it would be the closest thing we have to a panacea for economic development. However, we eﬀectively have a policy intervention that does this, which is to not impose these policies that signiﬁcantly harm developing countries. Thus, by removing the current practice, which is aimed at shifting rents from the less developed world, we could tangibly reduce poverty. By comparison, consider The Millennium Villages Project, which was a high proﬁle 10-year, multi- sector, rural development project that began in 2005 in 14 village sites in 10 countries (Mitchell et al. 2004). This project, with its multimillion dollar price tag, induced an improvement in income that pales in comparison to the beneﬁt that one would obtain by not initiating an antidumping petition against Catﬁsh farmers in Vietnam. Antidumping duties are but one example of actions that developed coun- tries take to the detriment of less-developed countries. However, there are many more examples. One that is closely related is tariﬀs in general. There is ample evidence that, like anti-dumping duties, tariﬀs placed against developing- country products increase poverty, reduce growth and retard industrialization 10 N. Nunn (e.g., McCaig 2011, McCaig and Pavcnik 2018). Despite this, tariﬀs have been, and continue to be, systematically higher against goods that developing coun- tries produce. They are higher in less-skilled industries for which developing countries have a comparative advantage (Nunn and Treﬂer 2013). Even within industries—i.e., at the product level—there is a bias against poor countries. Goods and varieties that are of lower quality and tend to be produced by less- developed countries, tend to have higher tariﬀs placed against them (Acosta and Cox 2019). Another example is the use of power and coercion in the international arena. Several papers have found that coercion has been used to the eco- nomic beneﬁt of those with power (e.g., developed nations) at the expense of those who do not have it (e.g., less-developed nations). Berger et al. (2013) show that CIA interventions that led to the instalment of “puppet leaders” who were aligned with the United States, whether through propaganda, election support, organized coups and assassinations, resulted in an increase in power that was used to create an export market for US products. After “puppet leaders” were installed, the sales of products from the US to the intervened country increased dramatically while the sales of the intervened country to the United States did not change. The increased sales seem to have been due to direct government purchases and that the goods purchased were ones that the United States was having a hard time ﬁnding a market for. It seems to me that we have ended up in a strange equilibrium. With one hand, the Western developed nations are taking actions that have obvious deleterious eﬀects on developing countries. These decisions increase poverty and cause the persistence of underdevelopment. This occurs through trade policies, like tariﬀs or anti-dumping duties, and through practices in inter- national relations and political economy. With the other hand, we are trying (or at least purport to be trying) to help developing countries through foreign aid, which we know often has unintended negative consequences and is typically given in a self-serving manner, e.g., through tied aid or aid given in kind. Probably the strategy that we as the West could be doing, is to not take these actions that are causing harm. That is, we don’t need to “ﬁx” anything, but we could stop harming developing countries. This raises the question of whether there are other policy options or things that can be done to reduce global poverty. I’ll raise a few possible policies here. The ﬁrst is greater labour mobility. There is now accumulating evidence showing that there are signiﬁcant beneﬁts to labour mobility. Not surprisingly, numerous studies have found that those who migrate beneﬁt. A recent study by Clemens et al. (2019) carefully accounts for selection into migration to obtain credible estimates of the causal eﬀects of migration. They ﬁnd that for unskilled male migrants with 9–12 years of education who move to the United States, migrating increases their annual real wage by 395%, or $13,600 dollars. This illustrates an unsurprising but important fact: there is a remarkable in- crease in well-being for those who migrate. Also, if we presume that individuals Rethinking economic development 11 are paid their marginal product, then this reﬂects a nearly 400% increase in the migrant’s productivity from moving to the United States. For many, it is unsurprising that migrants themselves are better oﬀ after relocating. However, how could greater labour mobility be the basis of a serious development strategy? Won’t all developing countries empty and countries that are currently wealthy become overcrowded? This line of thinking, which is common, ignores an important fact. This is that developing countries, including those left behind, are also made better oﬀ when someone migrants to another country. Evidence shows that members of the extended family who are left behind are also made better oﬀ due to remittances (Yang 2008, 2011), which themselves cause increased human capital accumulation, less child labour and more entrepreneurship and self-employment (Yang 2008). As Sanchez (2019) has documented, particularly successful migrants often ﬁnance public goods and other large-scale projects in the origin villages. Remittances are also sizable quantitatively. Despite the limited migration that occurs currently, they already total more than aggregate aid ﬂows (Yang 2011). Thus, the potential beneﬁts from migration due to remittances alone is large and with loosened migration restrictions remittances would easily swamp foreign aid ﬂows. It is not only the families that beneﬁt when an individual migrates. Evidence is accumulating suggesting sending countries can also beneﬁt from out-migration. Swedish emigration to the United States in the late 19th and early 20th centuries led to higher unionization rates, more civic participation and more inclusive domestic institutions (Karadja and Prawitz 2019) as well as greater innovation arising due to the resulting relative scarcity of labour (Andersson et al. 2017). Emigration to the United States also led to more long- term inward and outward FDI for the sending countries. Thus, international migration facilitated the creation of international business links that continue to exist today. Studies of more recent labour movements also ﬁnd similar eﬀects. Kerr (2008) ﬁnds that immigration to the United States results in greater knowledge ﬂows to the origin country, which results in greater productivity and output. On a more micro (and personal) level, I see many examples of this in our profession. One example is my co-author Leonard Wantchekon, who left Benin—or, more accurately, escaped from jail where he was imprisoned for pro-democracy activism—and ﬂed to Canada as a political refugee. In Canada, he obtained his MA in 1992 and then went to the United States, where he obtained his PhD in 1995. He has since gone on to be an extremely successful scholar, working in the ﬁelds of political economy and development economics.1 As is common amongst successful diaspora from developing coun- tries, he wanted to give back to his country. In 2014, he founded the African School of Economics, which is a PhD-granting university located in Cotonou, 1 For the full story, see Wantchekon (2012). 12 N. Nunn Ethiopia No data Visa-free Visa upon arrival Visa required FIGURE 4 Visa restrictions faced by Ethiopians when travelling to each country SOURCE: Umana-Dajud (2019) Benin. The university has already been extremely successful at educating and training many talented African-born students who have gone on to have fruitful careers in government or academia. Not only has the country of Benin beneﬁtting from Leonard’s migration but actually so has much of Africa. The last beneﬁts of migration that I will discuss are those for the receiving country. Sequeira et al. (2019) study immigration into the United States dur- ing its “age of mass migration.” They found that immigrants generated sizable economic beneﬁts in the locations in which they settled. These beneﬁts were felt almost immediately, persisted and continue to be felt today. Such eﬀects appear to be general and found in many diﬀerent settings. Examples range from the 17th-century Huguenot immigration into Prussia (Hornung 2014) to the large inﬂux of Vietnamese refugees into the United States beginning in the 1970s (Parsons and Vezina 2018) to immigration into France from 1995–2005 (Mitaritonna et al. 2017). While most studies estimate the eﬀects of immigration by looking at episodes of increased migration, Clemens et al. (2018) do this by studying the eﬀects of the elimination of the Mexican bracero worker agreement in 1964, a policy that resulted in the removal of nearly half a million seasonal Mexican farmworkers from the US labour force. They ﬁnd no evidence that domestic workers beneﬁtted from the exclusion, either in terms of higher wages or greater employment. Farm owners were forced to adjust to the scarcity of labour by shifting to labour-saving machines if such technology existed. If they could not do this, then they were forced to scale back production. While a dramatic loosening of immigration restrictions may not be feasible for political reasons, there are small changes that could be made in terms of labour restrictions that would have huge eﬀects. Take the example of restrictions and requirements on travel visas. Figure 4, which is taken from Umana-Dajud (2019), shows the visa restrictions facing Ethiopians who want to enter a foreign country. Each country is shaded according to restrictions faced. As can be seen, nearly every country in the world requires a visa to be obtained ahead of time. The handful of countries that allow a visa upon Rethinking economic development 13 arrival do not have direct ﬂights from Ethiopia. The transit countries all require a visa beforehand. Eﬀectively travel is very diﬃcult for any person from Ethiopia. This must aﬀect international business relations between Ethiopia and the rest of the world, especially given that it is now well documented that very small travel costs—i.e., the cost diﬀerence between a direct ﬂight and a two-leg ﬂight—have important eﬀects on business travel, capital ﬂows and economic growth (Campante and Yanagizawa-Drott 2018). Consistent with this, a recent study by Umana-Dajud (2019) ﬁnds that visa restrictions have a signiﬁcant negative eﬀect on trade ﬂows and are estimated to reduce the aggregate welfare of the most-aﬀected countries by as much as 5%. Another strategy to combat global poverty is the use of consumer purchas- ing power. There are a number of cases where consumers, by voting with their wallets, have been able to aﬀect change for the global poor. For example, the public outcry in the 1990s against Nike, Adidas and Reebok sourcing their products from sweatshop factories resulted in pressure that led to changes in production and labour standards in their factories. Harrison and Scorse (2010) document the eﬀects that this had for Indonesian workers. It resulted in higher wages, as the Indonesian government increased the minimum wage by 400%. At the same time, this does not appear to have come at the expense of less production or fewer jobs. Similarly, recent evidence by Dragusanu and Nunn (2018) ﬁnds that con- sumer purchases of fair trade-certiﬁed coﬀee result in increased incomes of FT-certiﬁed coﬀee farmers in Costa Rica. These beneﬁts ultimately originate due to consumers’ willingness to pay more for products that are produced and sourced in a manner that improves the lives of producers. Traditionally, we think of consumers as valuing only the attributes of the good they purchase. However, there is now overwhelming evidence that, above and beyond the product itself, consumers also value the production process and how it aﬀects the well-being of those who are involved in it. There have been numerous rigorous lab-in-the-ﬁeld experiments showing this. Typically, these involve randomly displaying information about a product’s certiﬁcation or not and also randomly changing its sales price. Such studies have examined fair trade-certiﬁed products (Arnot et al. 2006, Hiscox et al. 2011a, Hainmueller et al. 2015), products with fair labour standards (Hiscox and Smyth 2011, Hiscox et al. 2011b) and products with environ- mental standards (Hainmueller and Hiscox 2015). The general ﬁndings of the studies are that consumers are willing to pay signiﬁcantly more for a product that they believe to have been produced in a manner that improves the well-being of the producers who are in less- developed countries. In addition, with certiﬁcation, increasing the price does not decrease demand and in many cases further increases it. Thus, consumers often have the highest demand for a product when the price is high and it is known to be certiﬁed. Since the inception of certiﬁcation programs, the most well known being Fairtrade in 1997, sales of certiﬁed products has grown exponentially. This is 14 N. Nunn an indication of consumers untapped demand for the production of products that occurs in a way that beneﬁts workers in developing countries. As of 2016, the last year for which data are available, there are now over 1,400 organizations and 1.6 million farmers and workers, spanning 73 countries and 19 product categories that are Fairtrade certiﬁed (Fairtrade International 2018). This, combined with the growing popularity of such labelling initiatives, suggests the possibility of this being a tool that helps alleviate poverty in developing countries. My aim here has not been to provide a deﬁnitive answer to which policies work and which do not. Instead, it has been to point out that the discussion and narrative in the policy world, and the academic world too, tend to focus on one policy (i.e., foreign aid) among many (e.g., industrial policies, reduced labour mobility, reduced trade barriers, product labelling initiatives, etc.). Given the evidence that foreign aid is often self-serving for the donors, its beneﬁts to developing countries can often be disappointing and it can have unintended negative consequences, I think it is fruitful to think more broadly about development policy than we currently do. 3. Rethinking academic research I now turn to the area of academic research. Although there is diversity in the type and style of research that is done in economic development, the typical strategy follows these three steps: (1) an intervention is identiﬁed that is believed to have eﬀects that are beneﬁcial for economic development or well-being, (2) the intervention is randomly implemented (if possible) and rigorously evaluated and (3) if successful, then the policy is scaled up and/or replicated elsewhere. In practice, step 1 is most easily accomplished if one can ﬁnd an organization that is implementing an intervention and then convince them to randomize treatment or the timing of rollout. Step 2 is the core of the research process and step 3 is often left for policy-makers or for follow-up research. A well-known paper and research agenda that follows this recipe is Miguel and Kremer (2004). While all children eventually received deworming pills during the campaign, its roll-out was done in a staggered manner that was randomized over three years. Large beneﬁts were found. The program was then scaled up nationally and then replicated in other countries around the world (see, e.g., Croke and Atun 2019). The knowledge learned from the research undertaken by Miguel and Kremer (2004) had an important eﬀect on subsequent policy decisions and aﬀected the well-being of millions of children around the world. This line of research has several characteristics, some of which are fairly unique within the ﬁeld of economics. First, there are signiﬁcant ﬁxed costs associated solely with the logistics of the intervention and data collection. The interventions themselves are typically large in scale, costly and logistically diﬃcult. In many cases, the implementation is done by a partner organization, Rethinking economic development 15 such as an NGO. However, even in these cases, simply evaluating an existing project is still very costly and very diﬃcult. Second, there is generally an implicit presumption of fairly universal re- lationships across time and space.2 If one truly believes that every case was completely diﬀerent, then the strategy of scaling up and trying successful interventions in other locations would be ﬂawed. The presumption is strongly imposed in the profession through the refereeing and publication process. During my time as an editor at the Journal of Development Economics, there have been many occasions where a referee has rejected a paper with credible identiﬁcation and on an important topic because the same (or similar) relationship had been examined in a completely diﬀerent part of the world. For example, “We already know the relationship between transportation infra- structure (e.g., roads or railways) and poverty because paper X has estimated the relationship in county Z. Thus, there is no added contribution of paper Y.” Perhaps ironically, referees also tend to view it as a deﬁciency if a paper obtains a relationship that is diﬀerent from a previous paper examining a diﬀerent part of the world. Third, the methods that one uses in diﬀerent countries and when studying diﬀerent questions are the same and are fully transportable across diﬀerent contexts. This consistency is an attractive feature that, in theory, allows for the implementation of best practices and for the comparability of estimates from diﬀerent environments. Fourth, as a consequence of the previous two characteristics, in-depth knowledge of the local context is not completely necessary, which allows researchers to have projects in a large number of diﬀerent locations. If causal relationships are similar across locations and we are using the best-practice statistical tools in all locations, this means that a researcher’s experience studying education in India, for example, would be suﬃcient for studying similar topics in Tanzania. These assumptions mean that development economists can produce research published in the ﬁeld’s very best journals from more locations compared to other ﬁelds that also ask similar questions about economic development and involve ﬁeldwork, such as anthropology. These characteristics of this line research have consequences for the way the literature evolves. I would argue that they contribute to three features of the ﬁeld of development economics currently.3 The ﬁrst is that research tends 2 Of course, this is not speciﬁc to RCTs or development economics but is a general tendency within quantitative research. 3 An additional consequence of these characteristics of research is that focus has shifted towards the study of individuals and households and towards questions relevant to them. At the same time, there has been a shift away from the study of ﬁrms and issues important for them. The reason is simple. It is logistically easier and less costly to study households than it is to study ﬁrms. As an example, consider the study by Bloom et al. (2013), which examined the eﬀects of management consulting on Indian textile plants. The full cost/value of the 16 N. Nunn to be focused on a very limited number of countries for which the necessary research infrastructure has been established. With existing infrastructure, a new researcher does not have to bear the ﬁxed costs that are required when working in a new country. At the extreme, one can simply hire a survey company that already exists in the location. It is hard to overestimate the costs and logistics needed to begin large-scale interventions and surveys in a new location without existing infrastructure. This is something that my co- authors (Sara Lowes, James Robinson and Jonathan Weigel) and I (perhaps foolishly) undertook when we began working in the DRC in 2013 to study the eﬀects of traditional state formation among the Kuba and Lele (Lowes et al. 2017). There was no real foreign presence in the largest city of the region, Kananga, nor was there any running water, sewerage or electricity. Setting up to undertake research in this environment required huge upfront investments. These ranged from the more obvious like obtaining permissions from local oﬃcials at multiple levels to training enumerators. However, it also involved a large amount of tedious and mundane tasks like ﬁnding houses and oﬃces; hiring a cook; buying pots, pans, wash basins, 55-gallon drums to store water, beds, mattresses, etc.; ﬁguring out how to purify water; and ﬁguring out how to generate a reliable source of electricity. There was a lot of learning-by-doing and so the ﬁrst attempts at the tasks were not always successful. The second characteristic is that, relative to other ﬁelds and what (in my mind) is optimal, the local context and setting are seldom taken into account in suﬃciently meaningful ways, either in terms of the research design or the interpretations of the ﬁndings. Closely related to this is the third characteristic, which is that, during the process of research, local scholars are seldom involved in any meaningful way, such as being co-authors on the project. As a concrete example of this, consider the Abdul Latif Jameel Poverty Action Lab (J-PAL). The organization has been at the forefront of academic global outreach. In terms of involving those from developing countries in academic research, the organization is as inclusive as it gets. Despite this, as of the time this article is being written, according to their website, of J-PAL’s network of 170 aﬃliated professors from 58 universities, there is only one university that is not from a country that is either European or a European oﬀshoot. This is the Indian Institute of Management. They have one aﬃliate from an African university, but this is the University of Cape Town, located in South Africa. The fact that research seldom uses the details of the local context in any meaningful way goes a long way in explaining why local scholars are intervention that was implemented was $250,000 per observation (plant). This is much more than the cost per observation in a typical study that occurs at the household level. The researchers were able to overcome this by having a modest number of treated observations (only 14 in total) and by obtaining a signiﬁcant discount from the ﬁrm implementing the management consulting. Rethinking economic development 17 seldom involved in projects. If it is perceived that a deep understanding of the local context is not needed, then this signiﬁcantly reduces the incen- tives to reach out to local scholars and involve them in the research process. However, I would contend that this view is incorrect. Undertaking research that makes progress on development issues isn’t possible without a deep understanding of the local context. Without this, at best, our understanding is incomplete, and, at worst, it is completely oﬀ the mark. To illustrate this point, in subsection 3.1, I provide three examples of research to illustrate the point that understanding the local context is crucial for development research. The examples I have chosen are just a few of many. I have made an eﬀort to choose examples that are of the standard development poli- cies; namely, health and medicine, agriculture and education. I turn to these now. 3.1. Health and medical policies One of the areas in which foreign aid has the largest scope to improve the well- being of the global poor is in the area of health. Over the past century, there have been enormous medical advances. We now have vaccines and treatments that prevent and cure diseases that would otherwise be deadly, such as polio, malaria, yellow fever, trypanosomiasis, etc. Despite this, there are numerous examples of populations in developing countries refusing life-saving treatment. One of the more dramatic examples is of individuals refusing treatment in the 2014–2016 Ebola outbreak in West Africa and during the current outbreak in the Eastern DRC. This is a remarkable situation. We have made medical advances. We know how to save lives. Funds have been raised and resources mobilized. But, at the very last step of the process, when the medical treatment is oﬀered to recipients, they refuse. A natural knee-jerk reaction is to view this behaviour as irrational and inexplicable. However, the behaviour becomes understandable when one learns about the history of medical interventions on the African continent. The recent study by Lowes and Montero (2017) examines medical interventions that occurred during the colonial period in French Equatorial Africa and were aimed at eradicating sleeping sickness (trypanosomiasis). The medical campaigns required villagers, often at gunpoint, to submit to physical exams and tests for sleeping sickness. The only eﬀective test at the time was a spinal tap, an extremely painful procedure. Given this, it was often easier to simply treat whole villages, rather than only those with the disease. The early treatment used was an arsenic-based drug called atoxyl, which, it was later discovered, had the side eﬀect of causing at least partial blindness in approximately 20% of those treated. This dramatic experience with Western medicine continues to live in the collective memory of those alive today. The parents, grandparents and great grandparents of those alive today directly experienced the medical 18 N. Nunn interventions, which continued regularly until the end of colonial rule. Stories are told and songs sung of these traumatic historical experiences. Could these past medical interventions explain why individual’s today are hesitant to put their faith in Western medicine brought in by the de- veloped West? The study by Lowes and Montero (2017) seeks to answer this question. To do this, the authors went to archives in France to col- lect and digitize information on the frequency with which diﬀerent loca- tions within French Equatorial Africa were visited by the colonial medical campaigns. With this information, they are able to compare how the his- torical experience with colonial medical campaigns is associated with the success of contemporary foreign medical interventions. They ﬁnd that places with more visits in the past are also the places where medical interven- tions are less successful today. Interestingly, this negative eﬀect is found only for medical projects and not for other projects, such as infrastructure. This suggests that the adverse eﬀects have not spilled into other forms of foreign aid. Digging deeper into mechanisms, they also show that refusal rates for blood tests (for anemia and HIV/AIDS) and vaccines are signiﬁ- cantly higher in locations that experienced more colonial medical campaigns historically. This example illustrates the importance of understanding the historical and current cultural context of a location. Once these are known, behaviour that the West often views as irrational and frustrating becomes understandable. In addition, the potential solutions become more clear. The history of these locations suggests that much more needs to be done besides simply showing up and asking individuals to line up for vaccinations or blood tests. Instead, it is likely that signiﬁcant outreach, communication and a sensitivity to past experiences with Western medicine is essential. 3.2. Agricultural policies An example of the importance of local culture and context within agriculture is documented in the book Priests and Programmers by J. Stephen Lansing (1991). He describes the Hindu-based belief system in Bali called Agama Tirtha. It is a system where water is central and is viewed as holy. It features an elaborate hierarchical system of temples that are placed at the critical junctures of a complex irrigation system, which brings water to ﬁelds that are used for wet rice cultivation. The water originates from a volcanic crater lake that is located in the mountain of the island and where the water goddess, Dewi Danu, lives. Farmers in each rice terrace are organized into local democratic collectives called subak. Using religious calendars and rituals, all subaks in an area co- ordinate planting with each other, performing elaborate multiday rituals before planting. These involve a pilgrimage to the crater lake, then rituals at the regional water temples, followed by rituals at village temples and then rituals in the ﬁelds themselves. Rethinking economic development 19 In the mid-1970s, the Green Revolution was brought to the island. As part of this, farmers were encouraged to abandon old traditions and adopt new methods to increase their yields. Part of these eﬀorts was getting farmers to decouple their planting decisions with their spiritual rituals. They could continue these, but they should no longer wait and coordinate planting with other farmers in the region. This way farmers could plant as many times during the year as possible and not have to wait for their neighbours to do so. Within years of implementation, crops were consistently overwhelmed with insects. It turns out that the spiritual beliefs and religious rituals that co- incided with the planting decisions were beneﬁcial and had served to control pest populations. Because of the coordinated planting, all land within an area was fallow at the same time. In addition, for spiritual reasons, fallow land was ﬂooded. An unknown consequence of this is that it helped control pests. For a speciﬁc period of each year, the insects had nothing to eat and died oﬀ. By forcing farmers to abandon their traditional, religion-based farming practices, they unintentionally forced farmers into inferior farming practices. In the late 1980s, after much damage had been done, farmers returned to their traditional farming practices. In this example, insuﬃcient understanding of the local customs and cul- tural context led to policy interventions that did more harm than good. Despite being armed with superior agricultural technologies, due to the use of a one-size-ﬁts-all policy these interventions did more harm than good in Bali. 3.3. Education policies In 1973, the Indonesian government launched a program that built a remark- able 61,807 primary schools over seven years. It was the largest such school construction program ever undertaken. As is documented by Duﬂo (2001), the program had large positive eﬀects on the education of boys, which in turn resulted in higher incomes in their adulthood. In other words, the program was a success. Duﬂo’s well-known study looked at only boys because her primary interest was in estimating the causal eﬀect of education on adult wages. Since formal employment is low among women, they were not included in her study. How- ever, if one uses the same methodology, but examines girls rather than boys, one ﬁnds no eﬀect of the program on girls education (Ashraf et al. 2019). As Ashraf et al. (2019) show, this zero eﬀect masks important heterogeneity. For some ethnic groups, the eﬀect of the program was positive and very large. For others, the eﬀect was zero. After showing this, Ashraf et al. (2019) show that a similar pattern is found for another large-scale school construction program, in Zambia, in the late 1990s and early 2000s. It turns out that the key diﬀerence between the two groups, in each country, is that one comprises ethnic groups that have a custom of bride 20 N. Nunn price while the other comprises ethnic groups that do not. Bride price, which is also known as bridewealth, is a transfer of money and/or goods from the groom to the bride’s parents at the time of marriage. The practice is widespread in Asia and Africa and the payments are sizable, often more than one or two years’ wages. Among ethnic groups that practice bride price, schooling construction caused an increase in female educational attainment. For those that did not, school construction had no eﬀect. The connection between the practice and female education is not immediately obvious. Ashraf et al. (2019) provide the answer, which is due to an important fact that is well known to the local populations. This is that the value of the bride price—i.e., the amount that parents receive when their daughters are married—is signiﬁcantly higher the more education their daughter has. The authors document that completing primary school almost doubles the value of the bride price and completing secondary school almost doubles it again. Thus, the monetary returns to parents from their daughters are higher in societies with a bride price than those without. The authors develop and test a model that shows how the additional monetary beneﬁt from daughters’ education causes the eﬀectiveness of education policies, like school construction, to be much greater when the custom of bride price is present. This ﬁnding is particularly interesting given the common view by the de- veloped West of bride price as a primitive custom that involves the purchase of a wife by the husband. Partly motivated by this view, there have been numerous calls for its abolishment. The perception by those who practice the custom is not that the wife is being purchased by the husband, but that the bride price is a gift of appreciation to the parents for raising their daughter even knowing that she will be married at which time she will no longer be part of the parent’s family and will instead be a part of the husband’s family. Finding causal estimates of the eﬀect of bride price on the quality of the marriage and female empowerment is diﬃcult. However, existing correlations suggest that, conditioning on observables, the size of the bride price is positively associated with the quality of the marriage, female empowerment and the wife’s happiness (Lowes and Nunn 2018). The example shows how the success of a standard policy, like building schools to increase the educational attainment of women, depends critically on the customs of the population in question. Even within the same country, among some groups the policy is a huge success, but among others it is a complete failure. In addition, the contextual detail, in this case a marriage custom, is not something that would immediately come to mind when thinking about what customs are important for child education. This highlights the importance of a holistic understanding of a society and its cultural traits when trying to assess the eﬀectiveness of policies. Simply coming in and asking the questions that a researcher feels are relevant for education policy may not be enough. Intimate local knowledge is needed just to know what aspects of life are important for education. Rethinking economic development 21 4. Conclusions In this article, I have reﬂected on the current state of economic development, in both the policy world and in academia. As I have discussed, development policy tends to focus on interventions, typically funded with foreign aid, that are aimed at ﬁxing deﬁciencies in developing countries. The general perception is that there are inherent problems with less-developed countries that can be ﬁxed with the help of the Western world. I have discussed evidence that shows that the eﬀects of this “help” can be mixed. While there are beneﬁts, there can also be unintended adverse consequences. At the same time that this “help” is being oﬀered, the developed West undertakes many actions that are harmful to developing countries in obvious ways. Examples include tariﬀs, antidumping duties, limits on international labour mobility, the use of international power and coercion and tied aid used for export promotion. These are but a few examples. Thus, it is unclear whether interactions with the West are, on the whole, helpful or detrimental to developing countries. We may have our largest and most positive eﬀects on alleviating global poverty if we focused on restraining ourselves from actively harming less-developed countries rather than focusing our eﬀorts on ﬁxing them. Academic research mirrors the policy world in its general assumption that there are deﬁciencies in developing countries that help and insights from Western academics can help to solve. An important diﬀerence between the academic world and the policy world is that the former places much more emphasis on rigorous evaluation of interventions and the estimation of credible causal impacts. The set of methods that are used are solely quantitative and can be applied to essentially any setting. There is also usually a strong prior that the eﬀects being estimated are similar and informative across settings. Due to the ﬁxed costs of setting up the infrastructure needed for large-scale quantitative research, most research takes place in a limited set of countries with the necessary infrastructure. Since relationships are viewed as being similar across settings, this is not generally viewed as an issue. Relationships can be estimated in locations that are logistically easier to work in and we can extrapolate to locations that are more diﬃcult to work in. However, an upshot of this strategy is that we have come to under-appreciate the diﬀerences across contexts and their importance for research. Local context is not believed to be that important. As a consequence of this, local knowledge and local scholars are seldom involved in the research process in any meaningful way, e.g., as collaborators and co-authors. I ended by providing numerous examples of how this one-size-ﬁts-all view of policy and research can result in either an incomplete or incorrect understand- ing of the consequences of development policy. In these settings, a detailed understanding of the local context is necessary for proper analysis and eﬀective policy. However, this requires a number of things that are currently missing from development research, including research that incorporates qualitative 22 N. Nunn research (e.g., interviews, focus groups, etc.) in a real way, cross-disciplinary communication with anthropologists and historians and, most importantly, collaboration with local researchers. Once the academic world realizes the importance of local context for development policy, then the importance of local researchers becomes clear, which will go a long way towards making research more inclusive. References Acosta, M., and L. Cox (2019) “The regressive nature of the U.S. tariﬀ code: Origins and implications,” working paper, Columbia University Ahmed, F. Z. (2016) “Does foreign aid harm political rights? Evidence from U.S. aid,” Quarterly Journal of Political Science 11(2), 118–217 Alesina, A., and D. Dollar (2000) “Who gives aid to whom and why?,” Journal of Economic Growth 5(1), 33–63 Anderson, M. B. (1999) Do No Harm: How Aid Can Support Peace – Or War, London: Lynne Rienner Publishers Andersson, D., M. Karadja, and E. Prawitz (2017) “Mass migration, cheap labor, and innovation,” working paper, Institute for International Economic Studies, Stockholm University Arnot, C., P. C. Boxall, and S. B. Cash (2006) “Do ethical consumers care about price? A revealed preference analysis of fair trade coﬀee purchases,” Canadian Journal of Agricultural Economics 54(4), 555–65 Ashraf, N., N. Bau, N. Nunn, and A. Voena (2019) “Bride price and female education,” Journal of Political Economy, forthcoming Baird, S., J. H. Hicks, M. Kremer, and E. Miguel (2016) “Worms at work: Long-run impacts of a child health investment,” The Quarterly Journal of Economics 131(4), 1637–80 Barnett, M. (2011) Empire of Humanity: A History of Humanitarianism. Ithaca, NY: Cornell University Press Berger, D., W. Easterly, N. Nunn, and S. Satyanath (2013) “Commercial imperialism? Political inﬂuence and trade during the cold war,” The American Economic Review 103(2), 863–96 Bhagwati, J. N. (1982) “Directly unproductive, proﬁt-seeking (DUP) activities,” Journal of Political Economy 90, 988–1002 Bloom, N., B. Eifert, A. Mahajan, D. McKenzie, and J. Roberts (2013) “Does management matter? Evidence from India,” The Quarterly Journal of Economics 128(1), 1–51 Brambilla, I., G. Porto, and A. Tarozzi (2012) “Adjusting to trade policy: Evidence from US antidumping duties on Vietnamese catﬁsh,” The Review of Economics and Statistics 94(1), 304–19 Campante, F., and D. Yanagizawa-Drott (2018) “Long-range growth: Economic development in the global network of air links,” The Quarterly Journal of Economics 133(3), 1395–458 Clay, E. J., M. Geddes, and L. Natali (2009) “Aid untying: Is it working? An evaluation of the implementation of the Paris Declaration and of the 2001 DAC recommendation of untying ODA to the IDCs,” Danish Institute for International Studies Rethinking economic development 23 Clemens, M. A., E. G. Lewis, and H. M. Postel (2018) “Immigration restrictions as active labor market policy: Evidence from the Mexican bracero exclusion,” The American Economic Review 108(6), 1468–87 Clemens, M. A., C. E. Montenegro, and L. Pritchett (2019) “The place premium: Bounding the price equivalent of migration barriers,” The Review of Economics and Statistics 101(2), 201–13 Croke, K., and R. Atun (2019) “The long run impact of early childhood deworming on numeracy and literacy: Evidence from Uganda,” PLOS Neglected Tropical Diseases 13(1), e0007085 Crost, B., J. Felter, and P. B. Johnson (2014) “Aid under ﬁre: Development projects and civil conﬂict,” The American Economic Review 104(6), 1833– 56 (2016) “Conditional cash transfers, civil conﬂict and insurgent inﬂuence: Experimental evidence from the Philippines,” Journal of Development Economics 118(1), 171–82 Dragusanu, R., and N. Nunn (2018) “The impacts of fair trade certiﬁcation: Evidence from coﬀee producers in Costa Rica,” working paper, Harvard University Dube, O., and S. Naidu (2015) “Bases, bullets, and ballots: The eﬀect of US military aid on political conﬂict in Colombia,” Journal of Politics 77(1), 249–67 Duﬂo, E. (2001) “Schooling and labor market consequences of school construction in Indonesia: Evidence from an unusual policy experiment,” The American Economic Review 91(4), 795–813 Fairtrade International (2018) “Monitoring the scope and beneﬁts of fairtrade,” monitoring report, 9th edition. Available at tinyurl.com/fairtrade-org-uk Giorcelli, M. (2019) “The long-term eﬀects of management and technology transfers,” The American Economic Review 109(1), 1–33 Hainmueller, J., and M. J. Hiscox (2015) “The socially conscious consumer? Field experimental tests of consumer support for fair labor standards,” working paper, MIT. Available at scholar.harvard.edu/files/hiscox/files/ socially_conscious_consumer.pdf Hainmueller, J., M. J. Hiscox, and S. Sequeira (2015) “Consumer demand for fair trade: Evidence from a multistore ﬁeld experiment,” The Review of Economics and Statistics 97(2), 242–56 Harris, R., I. Keay, and F. Lewis (2015) “Protecting infant industries: Canadian manufacturing and the National Policy, 1870–1913,” Explorations in Economic History 56(1), 15–31 Harrison, A., and J. Scorse (2010) “Multinationals and anti-sweatshop activism,” The American Economic Review 100(1), 247–73 Haushofer, J., J. Reisinger, and J. Shapiro (2015) “Your gain is my pain: Negative psychological externalities of cash transfers,” working paper, Princeton University Haushofer, J., and J. Shapiro (2016) “The short-term impact of unconditional cash transfers to the poor: Experimental evidence from Kenya,” The Quarterly Journal of Economics 131(1), 1973–2042 (2018) “The long-term impact of unconditional cash transfers: Experimental evidence from Kenya,” working paper, Princeton University 24 N. Nunn Hausman, R., J. Hwang, and D. Rodrik (2007) “What you export matters,” Journal of Economic Growth 12(1), 1–25 Hiscox, M. J., M. Broukhim, and C. S. Litwin (2011a) “Consumer demand for fair trade: New evidence from a ﬁeld experiment using eBay auctions of fresh roasted coﬀee,” working paper, Harvard University Hiscox, M. J., M. Broukhim, C. S. Litwin, and A. Woloski (2011b) “Consumer demand for fair labor standards: Evidence from a ﬁeld experiment on eBay,” working paper, Harvard University Hiscox, M. J., and N. F. Smyth (2011) “Is there consumer demand for fair labor standards? Evidence from a ﬁeld experiment,” working paper, Harvard University Hornung, E. (2014) “Immigration and the diﬀusion of technology: The Huguenot diaspora in Prussia,” The American Economic Review 104(1), 84–122 Juhasz, R. (2018) “Temporary protection and technology adoption: Evidence from the Napoleon blockade,” The American Economic Review 108(11), 3339–76 Karadja, M., and E. Prawitz (2019) “Exit, voice and political change: Evidence from Swedish mass migration to the United States,” Journal of Political Economy 127(4), 1864–925 Kerr, W. R. (2008) “Ethnic scientiﬁc communities and international technology diﬀusion,” The Review of Economics and Statistics 90(3), 518–37 Kono, D. Y., and G. R. Montinola (2009) “Does foreign aid support autocrats, democrats, or both?,” Journal of Politics 71(2), 704–18 Kuziemko, I., and E. Werker (2006) “How much is a seat on the Security Council worth? Foreign aid and bribery at the United Nations,” Journal of Political Economy 114(5), 905–30 Lane, N. (2017) “Manufacturing revolutions: Industrial policy and networks in South Korea,” working paper, IIES Stockholm University Lansing, J. S. (1991) Priests and Programmers: Technologies of Power in the Engineering Landscape of Bali. Princeton, NJ: Princeton University Press Lischer, S. K. (2005) Dangerous Sanctuaries: Refugee Camps, Civil War, and the Dilemmas of Humanitarian Aid. Ithaca, NY: Cornell University Press Lowes, S., and E. Montero (2017) “Mistrust in medicine: The legacy of colonial medicine campaigns in Central Africa,” working paper, Bocconi University Lowes, S., and N. Nunn (2018) “Bride price and the wellbeing of women.” In S. Anderson, L. Beaman, and J.-P. Platteau, eds., Towards Gender Equity in Development, pp. 117–38. Oxford, UK: Oxford University Press Lowes, S., N. Nunn, J. A. Robinson, and J. L. Weigel (2017) “The evolution of culture and institutions: Evidence from the Kuba kingdom,” Econometrica 85(4), 1065–91 McCaig, B. (2011) “Exporting out of poverty: Provincial poverty in Vietnam and U.S. market access,” Journal of International Economics 85(1), 102–13 McCaig, B., and N. Pavcnik (2018) “Export markets and labor allocation in a low-income country,” The American Economic Review 108(7), 1899–941 Miguel, E., and M. Kremer (2004) “Worms: Identifying impacts on education and health in the presence of treatment externalities,” Econometrica 72(1), 159–217 Mitaritonna, C., G. Oreﬁce, and G. Peri (2017) “Immigrants and ﬁrms’ outcomes: Evidence from France,” European Economic Review 96, 62–82 Rethinking economic development 25 Mitchell, S., A. Gelman, R. Ross, J. Chen, S. Bari, U. K. Huynh,... J. D. Sachs (2004) “The millennium villages project: A retrospective observational, endline evaluation,” The Lancet 6(e500–e513), 725–53 Mitrunen, M. (2019) “War reparations, structural change, and intergenerational mobility,” working paper, IIES Stockholm University Moscona, J. (2019) “The management of aid and conﬂict in Africa,” working paper, MIT Nunn, N., and N. Qian (2014) “U.S. food aid and civil conﬂict,” The American Economic Review 104(6), 1630–66 (2013) “The structure of tariﬀs and long-term growth,” American Economic Journal: Macroeconomics 2(4), 158–94 Olken, B. (2007) “Monitoring corruption: Evidence from a ﬁeld experiment in Indonesia,” Journal of Political Economy 115(2), 200–49 Parsons, C., and P.-L. Vezina (2018) “Migrant networks and trade: The Vietnamese boat people as a natural experiment,” The Economic Journal 128(612), F210–F234 Polman, L. (2010) The Crisis Caravan: What’s Wrong with Humanitarian Aid? New York: Henry Holt and Co. Prusa, T. J. (2001) “On the spread and impact of antidumping,” Canadian Journal of Economics 34(3), 591–611 Reinikka, R., and J. Svensson (2004) “Local capture: Evidence from a central government transfer program in Uganda,” Journal of Political Economy 119(2), 679–705 Sanchez, M. F. (2019) “Mass migration and education over a century: Evidence from the Galician diaspora,” working paper, Paris School of Economics Sequeira, S., N. Nunn, and N. Qian (2019) “Immigrants and the making of America,” The Review of Economic Studies, forthcoming Terry, F. (2002) Condemned to Repeat: The Paradox of Humanitarian Action. Ithaca, NY: Cornell University Press Trisko Darden, J. (2020) Aiding and Abetting: U.S. Foreign Assistance and State Violence. Stanford, CA: Stanford University Press Umana-Dajud, C. (2019) “Do visas hinder international trade in goods?,” Journal of Development Economics 140, 106–26 Wantchekon, L. (2012) Rêver à contre-courant: Autobiographie. Paris: L’Harmattan Werker, E., F. Z. Ahmed, and C. Cohen (2009) “How is foreign aid spent? Evidence from a natural experiment,” American Economic Journal: Macroeconomics 1(2), 225–44 Yang, D. (2008) “International migration, remittances and household investment: Evidence from Philippine migrants’ exchange rate shocks,” The Economic Journal 118, 591–630 (2011) “Migrant remittances,” Journal of Economic Perspectives 25(3), 129–52 Journal of Economic Literature 2013, 51(2), 325–369 http://dx.doi.org/10.1257/jel.51.2.325 How Deep Are the Roots of Economic Development? † Enrico Spolaore and Romain Wacziarg* The empirical literature on economic growth and development has moved from the study of proximate determinants to the analysis of ever deeper, more fundamental factors, rooted in long-term history. A growing body of new empirical work focuses on the measurement and estimation of the effects of historical variables on contemporary income by explicitly taking into account the ancestral composition of current populations. The evidence suggests that economic development is affected by traits that have been transmitted across generations over the very long run. This article surveys this new literature and provides a framework to discuss different channels through which intergenerationally transmitted characteristics may impact economic development, biologically (via genetic or epigenetic transmission) and culturally (via behavioral or symbolic transmission). An important issue is whether historically transmitted traits have affected development through their direct impact on productivity, or have operated indirectly as barriers to the diffusion of productivity- enhancing innovations across populations. (JEL J11, O33, O47, Z13) “The further backward you look, have evolved over time. Decades ago, the the further forward you can see” emphasis was on the accumulation of fac- (attributed to Winston Churchill).1 tors of production and exogenous techno- logical progress. Later, the focus switched 1. Introduction to policies and incentives endogenously affecting factor accumulation and innova- W hy is income per capita higher in some societies and much lower in others? Answers to this perennial question tion. More recently, the attention has moved to the institutional framework underly- ing these policies and incentives. Pushing * Spolaore: Tufts University, National Bureau of Eco- nomic Research, and CESIfo. Wacziarg: University of California at Los Angeles, National Bureau of Economic Research, and CEPR. We thank Leonardo Bursztyn, Janet 1 This is the usual form of the quote attributed to Currie, Oded Galor, David Weil, and several anonymous Winston Churchill—for instance, by Queen Elizabeth II referees for useful input. in her 1999 Christmas Message. According to Langworth † Go to http://dx.doi.org/10.1257/jel.51.2.325 to visit the (2008, 577), Churchill’s words were “the longer you can article page and view author disclosure statement(s). look back, the farther you can look forward.” 325 326 Journal of Economic Literature, Vol. LI (June 2013) back the debate one more degree, a key affecting income and productivity over question remains as to why the proximate the long run.2 We review the literature determinants of the wealth of nations vary on the legacy of geographic conditions in across countries. A burgeoning literature section 2. seeks to better understand the deep causes A major theme emerging from the recent of development, rooted in geography and literature is that key human characteristics history. affecting development are transmitted from As the empirical literature has moved one generation to the next within popula- from studying the proximate determi- tions over the long run, explaining why deep nants of growth and development to ana- historical factors still affect outcomes today. lyzing ever deeper, more fundamental A growing body of new empirical work has factors, important questions have arisen: focused on the measurement and estimation How much time persistence is there in of long-term effects of historical variables development outcomes? How far back in on contemporary income by explicitly tak- time should we go in order to understand ing into account the ancestral composition of contemporary economic development? current populations (Spolaore and Wacziarg Through what specific mechanisms do 2009; Putterman and Weil 2010; Comin, long-term geographic and historical fac- Easterly, and Gong 2010; Ashraf and Galor tors affect outcomes today? If economic 2013). We survey contributions to this new development has deep historical roots, literature in section 3. what is the scope for policy to affect the In section 4, we provide a general taxon- wealth of nations? This article discusses omy to discuss different channels through the current state of knowledge on these which inherited human characteristics may issues, focusing on recent empirical work impact economic development. Our discus- shedding light on the complex interactions sion builds on an extensive evolutionary lit- among geography, history, and compara- erature on the complex interactions among tive development. Throughout, we illus- genetic, epigenetic, and cultural transmis- trate the major milestones of the recent sion mechanisms, and on the coevolution of literature in a unified empirical frame- biological and cultural traits (Cavalli-Sforza work for understanding variation in eco- and Feldman 1981; Boyd and Richerson nomic development. 1985; Richerson and Boyd 2005; Jablonka Our starting point is the long-standing and Lamb 2005), as well as on a growing debate on geography and development. literature on cultural transmission and eco- There is no doubt that geographic factors, nomic outcomes (e.g., Bisin and Verdier such as latitude and climate, are highly cor- 2000, 2001; Tabellini 2008, 2009; Alesina, related with development, but the inter- Giuliano, and Nunn 2013). An important pretation of this correlation remains hotly issue is whether historically transmitted debated. While some of the effects of geog- characteristics affect economic development raphy may operate directly on current pro- through their direct impact on productivity, ductivity, there is mounting evidence that or operate indirectly as barriers to the diffu- much of the correlation operates through sion of technological and institutional inno- indirect mechanisms, i.e., through the his- vations across populations. torical effects of initial geographic condi- tions on the spatial distribution of human 2 For recent discussions of these issues from differ- characteristics, such as institutions, human ent perspectives, see Galor (2005, 2011) and Acemoglu, capital, social capital, and cultural traits, Johnson, and Robinson (2005). Spolaore and Wacziarg: How Deep Are the Roots of Economic Development? 327 2. Geography and Development Table 1, column 1 shows that a small set of geographic variables (absolute latitude, the 2.1 Long-Term Effects of Geography percentage of a country’s land area located in tropical climates, a landlocked coun- The hypothesis that geographic factors try dummy, an island country dummy) can affect productivity and economic develop- jointly account for 44 percent of contempo- ment has a long pedigree, going back to rary variation in log per capita income, with Machiavelli (1531), Montesquieu (1748), and quantitatively the largest effect coming from Marshall (1890). A vast empirical literature absolute latitude (excluding latitude causes has documented high correlations between the R 2 to fall to 0.29). This result captures current levels of income per capita and a the flavor of the above-cited literature docu- series of geographic and biological variables, menting a strong correlation between geog- such as climate and temperature (Myrdal raphy and income per capita. 1968; Kamarck 1976; Masters and McMillan While the correlation between geography 2001; Sachs 2001), the disease environment and development is well established, the (Bloom and Sachs 1998; Sachs, Mellinger, debate has centered around causal mecha- and Gallup 2001; Sachs and Malaney 2002), nisms. A number of prominent economists, natural resources (Sachs and Warner 2001), including Myrdal (1968), Kamarck (1976), and transportation conditions (Rappaport and Sachs and coauthors, argue that geo- and Sachs 2003). graphic factors have a direct, contemporane- In order to illustrate the main empiri- ous effect on productivity and development. cal findings of the contributions discussed In particular, Sachs (2001) claims that eco- herein, we punctuate this paper with our nomic underdevelopment in tropical coun- own empirical results based on a unified data tries can be partly explained by the current set, regression methodology and sample. negative effects of their location, which This analysis is not meant to be an exhaustive include two main ecological handicaps: low recapitulation of existing results, but simply agricultural productivity and a high burden of to illustrate some important milestones in diseases. Tropical soils are depleted by heavy the recent literature. We use, alternately, log rainfall, and crops are attacked by pests and per capita income in 2005 (from the Penn parasites that thrive in hot climates without World Tables version 6.3) as a measure of winter frosts (Masters and McMillan 2001). contemporary economic performance, and Warm climates also favor the transmission of population density in 1500 (from McEvedy tropical diseases borne by insects and bacte- and Jones 1978) as a measure of economic ria, with major effects on health and human performance in 1500, and regress these on a capital. In sum, according to this line of variety of proposed determinants of develop- research, geography has direct current effects ment, starting here with geographic factors.3 on productivity and income per capita. Other scholars, in contrast, claim that 3 As is well known, in the preindustrial, Malthusian geography affects development indirectly era population density is the appropriate measure of a through historical channels, such as the society’s economic performance since any technologi- effects of prehistoric geographic and biologi- cal improvement leads to increases in population rather than to increases in per capita income. For a theoretical cal conditions on the onset and spread of agri- and empirical analysis of the relationship between popu- culture and domestication (Diamond 1997; lation size, population density, and long-term growth in Olsson and Hibbs 2005), and the effects Malthusian times, see Kremer (1993). For in-depth discus- sions of this topic, see Galor (2005) and the recent contri- of crops and germs on the settlement of bution by Ashraf and Galor (2011a). European colonizers after 1500 (Engerman 328 Journal of Economic Literature, Vol. LI (June 2013) Table 1 Geography and Contemporary Development (Dependent variable: log per capita income, 2005; estimator: OLS) Whole Olsson–Hibbs Olsson–Hibbs Olsson–Hibbs Olsson–Hibbs Old World Sample: World samplea samplea samplea samplea only (1) (2) (3) (4) (5) (6) Absolute latitude 0.044 0.052 (6.645)*** (7.524)*** Percent land area in –0.049 0.209 –0.410 –0.650 –0.421 –0.448 the tropics (0.154) (0.660) (1.595) (2.252)** (1.641) (1.646) Landlocked dummy –0.742 –0.518 –0.499 –0.572 –0.505 –0.226 (4.375)*** (2.687)*** (2.487)** (2.622)** (2.523)** (1.160) Island dummy 0.643 0.306 0.920 0.560 0.952 1.306 (2.496)** (1.033) (3.479)*** (1.996)** (3.425)*** (4.504)*** Geographic conditions 0.706 0.768 0.780 (Olsson–Hibbs)b (6.931)*** (4.739)*** (5.167)*** Biological conditions 0.585 –0.074 0.086 (Olsson–Hibbs)c (4.759)*** (0.483) (0.581) Constant 7.703 7.354 8.745 8.958 8.741 8.438 (25.377)*** (25.360)*** (61.561)*** (58.200)*** (61.352)*** (60.049)*** Observations 155 102 102 102 102 83 Adjusted R2 0.440 0.546 0.521 0.449 0.516 0.641 Notes: a The Olsson and Hibbs sample excludes the neo-European countries (Australia, Canada, New Zealand, and the United States) and countries whose current income is based primarily on extractive wealth (Olsson and Hibbs 2005). b First principal component of number of annual or perennial wild grasses and number of domesticable big mam- mals (all variables from Olsson and Hibbs 2005) c First principal component of absolute latitude; climate suitability to agriculture; rate of East–West orientation; size of landmass in millions of sq km (all variables from Olsson and Hibbs 2005). Robust t statistics in parentheses. *** Significant at the 1 percent level. ** Significant at the 5 percent level. * Significant at the 10 percent level. and Sokoloff 1997 and 2002; Acemoglu, in 10,000 BC (the Neolithic Revolution). Johnson, and Robinson 2001, 2002; Easterly These advantages included the larger size and Levine 2003). of Eurasia, its initial biological conditions Diamond (1997) famously argues that the (the diversity of animals and plants avail- roots of comparative development lie in a able for domestication in prehistoric times), series of environmental advantages enjoyed and its East–West orientation, which facili- by the inhabitants of Eurasia at the transition tated the spread of agricultural innovations. from a hunter–gatherer economy to agricul- Building on these geographic advantages, tural and pastoral production, starting roughly Eurasia experienced a population explosion Spolaore and Wacziarg: How Deep Are the Roots of Economic Development? 329 and an earlier acceleration of technological Columns 3–5 add the two main Olsson– innovation, with long-term consequences Hibbs geographic variables, first separately for comparative development. According to and then jointly: a summary measure of bio- Diamond, the proximate determinants of logical conditions and a summary measure European economic and political success of geographic conditions.5 Both geographic (“guns, germs, and steel”) were therefore and biological conditions variables are highly the outcomes of deeper geographic advan- significant when entered separately. When tages that operated in prehistoric times. The entered jointly, the geographic conditions descendants of some Eurasian populations variable remains highly significant and the (Europeans), building on their Neolithic overall explanatory power of the regressors advantage, were able to use their technologi- remains large (52 percent). These empiri- cal lead (guns and steel) and their immunity cal results provide strong evidence in favor to old-world diseases (germs) to dominate of Diamond’s hypotheses, while suggesting other regions in modern times—including that the geographic component of the story regions that did not enjoy the original geo- is empirically more relevant than the bio- graphic advantages of Eurasia. logical component. Column 6 goes further in In order to test Diamond’s hypotheses, the attempt to control for the effect of post- Olsson and Hibbs (2005) provide an empiri- 1500 population movements, by restrict- cal analysis of the relation between initial ing the sample to the Old World (defined biogeographic endowments and contempo- as all countries minus the Americas and rary levels of development.4 They use several Oceania). The effect of geography now rises geographic and biological variables: the size to 64 percent—again highly consistent with of continents, their major directional axis Diamond’s idea that biogeographic condi- (extent of East–West orientation), climatic tions matter mostly in the Old World.6 factors, and initial biological conditions (the number of animals and plants suitable to domestication and cultivation at each location 12,000 years ago). We revisit their 5 These are the first principal components of the empirical results in columns 2 through 5 of above-listed factors. Since latitude is a component of the geographic conditions index, we exclude our measure of table 1. In order to reduce the effect of post- latitude as a separate regressor in the regressions that 1500 population movements, the Olsson– include geographic conditions. 6 Olsson and Hibbs also find that geographic variables Hibbs sample excludes the neo-European continue to be positively and significantly correlated with countries (Australia, Canada, New Zealand, income per capita when they control for measures of the and the United States), as well as countries political and institutional environment. They show that whose current income is based primarily on such political and institutional measures are positively correlated with geographic and biogeographic conditions, extractive wealth. Column 2 replicates the consistent with the idea that institutions could mediate the estimates of column 1 using this restricted link between geography and development. As they notice sample—the joint explanatory power of geo- (934), controlling for political–institutional variables raises well-known issues of endogeneity and reverse causality (for graphic variables rises to 55 percent, since instance, richer countries can have the resources and abil- the new sample excludes regions that are ity to build b

Rethinking Economic Development PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue