Ethnic Networks and Price Dispersion

Abstract

We offer and test a model linking ethnic networks to global price dispersion which predicts lower price dispersion as shared ethnic populations between countries rise, effects that may reverse at higher levels as network discipline breaks down. Using Chinese, Indian and Japanese data, we find that country pairs linked by the Chinese network have significantly lower mean price dispersion. A one standard deviation increase in the size of the Chinese coethnic network lowers price dispersion by 6-33%, an effect that reverses as the network gets large. No such evidence exists for the Indian or Japanese networks. the first to propose and test an arbitrage model that links price dispersion to Chinese, Indian and Japanese ethnic networks. 2 Our first contribution is a theoretical model that relates the size of a given ethnic network between country pairs (hereafter "coethnic network") to price dispersion. Networks create discipline among their members, through the threat of expulsion, and this reduces transactions costs and price dispersion. Initially, then, a larger coethnic network between locations lowers price dispersion between those same locations. As network size increases further, however, expulsion becomes more difficult and discipline may fall, reducing the effect of the network on price dispersion.Our second contribution is an empirical test of the model's predictions. We find evidence that the Chinese coethnic network reduces price dispersion significantly. A one standard deviation increase in the size of the Chinese coethnic network between location pairs is associated with a decrease in price dispersion of between 6% and 33%. For the Chinese network the effect reverses, and price dispersion increases, as the network becomes large. The Indian network, while of commensurate size, fails to show a negative association with price dispersion. The Japanese network is tiny by comparison and there is again no evidence that it decreases price dispersion.Finally, we offer other empirical contributions including panel estimation and an instrumental variable (IV) estimate of the effect of coethnic networks on price dispersion. Our IV estimates are consistent with, and are in fact stronger than, our standard estimates.There is a nascent literature on immigrants' influence on price levels and price dispersion to which our contribution can be usefully compared. Lach (2007), Cortes (2008) and Zachariadis (2011) all study the influence of immigration on city-level prices in (respectively) Israel, the USA, and a sample of cities in the USA and eleven other countries. 3 Grouping immigrants from different countries together, they find that a higher share of immigrants in a city's population is associated with lower goods prices and price levels in individual locations. This effect is the net outcome of a variety of supply and demand side forces at work on goods prices. Zachariadis (2012) estimates the influence of immigration on goods price dispersion across locations, measuring price dispersion as the log deviation of a locat...

A Disaggregated Analysis of Product Price Integration in the Southern African Development Community

“…Price dispersion may arise due to the presence of direct barriers such as distance or transport costs (Dumas, 1992;Engel & Rogers, 1996;Crucini et al, 2005a;Bergin & Glick, 2007;Anderson, Schaefer & Smith, 2013); cost differences between markets -arising, for example, through variation in factor prices; heterogeneity in market structures; differences in income levels, language and cultural differences; or variation in the density of ethnic networks (Aker et al, 2010;Anderson, Davies & Smith, 2012).…”

“…Furthermore, trade flows within the region are highly concentrated both in terms of product coverage -the majority of intra-SADC trade occurs in the form of trade in resource-based products (Gillson, 2012) -and the dominance of trade between the Southern African Customs Union (SACU) countries (Behar & Edwards, 2011). 2 Importantly, however, diagnoses of the level of market integration based on the volume of trade flows between countries can be misleading since trade volumes may be an endogenous outcome of market integration, and are also influenced by unrelated factors such as government expenditure, exchange rates, donor funding and rules of origin (Edwards & Rankin, 2012). The latter are particularly onerous in the SADC context and may indirectly affect trade flows in the region (Flatters & Kirk, 2003).…”

E pluribus unum? Political fractionalization and deviations from the law of one price in the USA

“…Using these shares, it is possible to calculate the probability that a randomly selected city-i voter and a randomly selected city-j voter will both have voted for the same candidate. This empirical measure, denoted P(residents-both-vote-sameway-16), corresponds directly to the network density measure in theoretical models such as the one outlined in Anderson et al (2016). 4 Higher values of P(residents-both-vote-same-way-16) indicate lower levels of cross-city fractionalization, so we expect the coefficient on this variable to be negative.…”

“…The second area of research concerns the effect of social fractionalization on trade costs. There is evidence that when a larger proportion of the inhabitants of two cities or regions are from a common ethnic group, then there is more inter-city trade and less inter-city price dispersion (Rauch, 2001;Rauch and Trindade, 2002;Wagner et al, 2002;Kumagai, 2007;Anderson et al, 2016). One way in which common ethnicity could reduce trade costs is through a shared language, and there is substantial evidence for the effect of a shared language on trade flows and price dispersion, both in international trade (Engel and Rogers, 2001;Parsley and Wei, 2001b;Melitz, 2008;Egger and Lassmann, 2015) and in intra-national trade (Sauter, 2012;Fielding et al, 2015).…”

“…In order to interpret these results, Anderson et al (2016) present a generic theoretical model in which a common culture reduces the cost of trade to individuals through lower contracting or search costs; it can also facilitate disciplinary mechanisms that increase trust and discourage cheating. Aggregate trade costs are not necessarily monotonically decreasing in the proportion of individuals in different locations sharing a common culture, because discipline may be more difficult to enforce in a larger cultural group, but the relationship is negative for a wide range of plausible coefficient values.…”