Non-parametric gravity as defined in this paper characterizes operational implications of spatial arbitrage in goods. Intuition that structural gravity is more general than previous parametric forms is validated. Non-parametric sufficient statistics for gains from trade and terms of trade are derived. Terms of trade in manufacturing 2000-2014 reveal China's early improvement of 25% followed by an overall decline of 42% driven by its almost fourfold rise in world share. US manufacturing terms of trade declined 20%. Trade elasticities that best parameterize non-parametric gravity are less than half those commonly used in counterfactual exercises.
Atkin, Costinot, and Fukui study the relationship between international trade and development in a model where countries differ in their capability, goods differ in their complexity, and capability growth is a function of a country's pattern of specialization. Theoretically, they show that it is possible for international trade to increase capability growth in all countries and, in turn, to push all countries up the development ladder. This occurs because: (i ) the average complexity of a country's industry mix raises its capability growth, and (ii ) foreign competition is tougher in less complex sectors for all countries. Empirically, Atkin, Costinot, and Fukui provide causal evidence consistent with (i ) using the entry of countries into the World Trade Organization as an instrumental variable for other countries' patterns of specialization. The opposite of (ii ), however, appears to hold in the data. Through the lens of their model, these two empirical observations imply dynamic welfare losses from trade that are small for the median country, but pervasive and large among a number of African countries.
This paper was distributed as Working Paper 29500, where an updated version may be available.
Limão and Xu model the implications of the classical ideas that larger markets allow for a finer division of labor and this division feeds back into larger market size. Market size affects specialization due to firm-level increasing returns to scale arising from fixed costs of adopting intermediate-intensive technologies. The impacts are magnified in general equilibrium by an endogenous multiplier--due to input-output linkages in a roundabout structure--and a selection effect due to heterogeneous fundamental productivity and entry costs. Market size expansions imply (i) larger real income gains than under fixed specialization; (ii) an increase in the aggregate variable cost share for intermediates and a decrease for labor; (iii) increased concentration; (iv) increased average productivity for survivors; and (v) an increase in the intermediate trade share. Limão and Xu derive similar results for intermediate productivity improvements. The effects in (ii)(v) are absent in a similar model with exogenous specialization. In a calibration to U.S. manufacturing in 1987-2007 the researchers isolate trade and intermediate productivity shocks and quantify their effects. Trade cost reductions increased effective market size by 7 log points (lp) and generated (i) a real income gain 1.4 times higher than under exogenous specialization; (ii) increases in the intermediate share in production and trade of 2 lp and a reduction in the labor share of value added of similar magnitude. Two counterfactuals highlight the importance of industrial and trade policy. First, a tax that induces firms to specialize increases real income; so the initial equilibrium is inefficient. Second, an increase in trade costs of 16 lp--similar to the recent trade war--reduces market size and real income substantially: almost half way to trade autarky.
This paper was distributed as Working Paper 28969, where an updated version may be available.
Heise, Pierce, Schaur, and Schott develop a model of international procurement and show that increases in the probability of trade war inhibit domestic buyers' use of long-term relationships to ensure the provision of high-quality inputs. The researchers introduce a method for identifying such "Japanese" importing in transaction-level trade data, and find that a policy change that lowered the likelihood of trade war fostered a relative shift towards "Japanese" procurement among U.S. importers and Chinese exporters. Embedding their model in a standard multi-country general equilibrium model of Ricardian comparative advantage, Heise, Pierce, Schaur, and Schott show via quantitative simulations that such shifts can raise welfare appreciably by allowing buyers to avoid costly inspection.
This paper develops a new method for estimating production-function parameters that can be applied in differentiated-product industries with endogenous quality and variety choice. de Roux, Eslava, Franco, and Verhoogen take advantage of data on physical quantities of outputs and inputs from the Colombian manufacturing survey, focusing on producers of rubber and plastic products. Assuming constant elasticities of substitution of outputs and inputs within firms, they aggregate from the firm-product to the firm level and show how quality and variety choices may bias standard estimators. Using real exchange rates and variation in the "bite" of the national minimum wage, the researchers construct external instruments for materials and labor choices. de Roux, Eslava, Franco, and Verhoogen implement a simple two-step instrumental variables method, first estimating a difference equation to recover the materials and labor coefficients and then estimating a levels equation to recover the capital coefficient. Under the assumption that the instruments are uncorrelated with firms' quality and variety choices, this method yields consistent estimates, free of the quality and variety biases the researchers have identified. Their point estimates differ from those of existing methods and changes in their preferred productivity estimator perform relatively well in predicting future export growth.
This paper was distributed as Working Paper 28323, where an updated version may be available.
Free trade or preferential trade areas (PTAs) allow importers who belong to the area to export to each other while paying zero or preferential tariffs as long as Rules of Origin (ROOs) are met. Meeting them is costly not only in terms of production costs but also in terms of documentation costs. Krishna, Salamanca, Suzuki, and Volpe Martincus ask if these fixed costs of documentation change over time with the experience of the firm in obtaining preferential tariffs. Krishna, Salamanca, Suzuki, and Volpe Martincus explore this using a unique importer-exporter matched transaction-level customs data set on a group of Latin American countries. Their estimating equation is model-based and shows that these fixed costs depend on the history of preference utilization. Most of the effect comes from experience in the same product and same partner, with some spillover to other partners buying the same product. There is little learning from experience in other products and other partners. When considering products that have been under preferences for a while, some learning might have occurred prior to the start of their data. Using a natural experiment in Argentina, where some products were newly brought under preferences, Krishna, Salamanca, Suzuki, and Volpe Martincus show that learning is indeed larger for such products. As facilitating preference use today also makes it easier to use preferences in the future, interventions early on in the life of the FTA to reduce such costs would be more effective.
This paper was distributed as Working Paper 29319, where an updated version may be available.
Alessandria, Khan, Khederlarian, Ruhl, and Steinberg study the growth of U.S. imports from China, from autarky during 1950--1970 to 15 percent of overall imports in 2008, taking advantage of the rich heterogeneity in trade policy and trade growth across products during this period. Central to their analysis is an accounting for the dynamics of trade, trade policy, and trade-policy expectations: the researchers isolate the effects of uncertainty about future reforms from the lagged effects of past reforms. First, they document that the conventional measure of U.S.-China trade policy uncertainty---the amount that tariffs would have risen if China's Normal Trade Relations (NTR) status was revoked---also captures the lagged effects of the large trade liberalization that occurred when this status was first granted in 1980. Second, Alessandria, Khan, Khederlarian, Ruhl, and Steinberg estimate a path of trade-policy expectations using a multi-industry, heterogeneous-firm model with a dynamic export participation decision. The researchers find that being granted NTR status in 1980 was largely a surprise and that this reform initially had a high probability of being reversed. The likelihood of reversal dropped considerably during the mid 1980s, however, and changed little throughout the late 1990s and early 2000s despite China's accession to the World Trade Organization (WTO) in 2001.
Alfaro, Bao, Chen, Hong, and Steinwender investigate how firms adapt to trademark protection, a most extensively used but under-examined form of IP protection, by exploring a historical precedent: China's trademark law of 1923---an unanticipated, disapproved response to end conflicts between foreign powers. Exploiting a unique, newly digitized firm-employee-level dataset from Shanghai in 1870-1941, the researchers show that the trademark law impacted firm dynamics on all sides of trademark conflicts. The law spurred growth and brand investment for Western firms with greater dependence on trademark protection. In contrast, Japanese businesses, who had frequently been accused of counterfeiting, experienced contractions while attempting to build their own brands after the law. Further, the trademark law led to new linkages with domestic agents both within and outside the boundary of Western firms and the growth of Chinese intermediaries. At the aggregate level, trademark-intensive industries witnessed a net growth in employment and product categories. A comparison with previous attempts by foreign powers to strengthen trademark protection--such as extraterritorial rights, bilateral treaties, and an unenforced trademark code--shows the alternative institutions were ultimately unsuccessful.
Heblich, Redding, and Zylberberg provide new theory and evidence on the distributional consequences of trade using the 1846 Repeal of the Corn Laws. This large-scale trade liberalization opened domestic markets to the "grain invasion" from the new world that occurred as a result of late-19th century improvements in transport technology. They make use of a newly-created, spatially-disaggregated dataset on population, employment by sector, rateable values (land and property values), and poor law (welfare transfers) disbursement for around 11,000 parishes in England and Wales from 1801-1911. The researchers show that the repeal of the Corn Laws led to rural outmigration, increased urbanization, structural transformation away from agriculture, increases in rural poverty, and sizable changes in property values. Heblich, Redding, and Zylberberg show that a quantitative spatial model is successful in accounting for these empirical findings, with their estimates implying substantial labor mobility. The researchers find that the aggregate welfare gains from the Repeal of the Corn Laws entailed considerable income redistribution, not only across sectors and factors, but also across geographical regions.
Gervais, Markusen, and Venables explore the idea that the comparative advantage of regions within a country is shaped by their productivity in supplying 'functions' such as law, finance, advertising and engineering, to multiple sectors. The paper addresses two questions. How do region-function specific productivity differences shape the location decisions of industries that use multiple functions, and hence determine patterns of regional specialization both in functions and in sectors? How do changes in the ease with which industries can draw on functions produced in other regions affect these patterns of specialization? We derive theoretical answers to these questions in a model in which region-function specific productivity differentials may be exogenous or driven by agglomeration economies. The model's prediction that falling barriers to interregional trade in functions lead to increasing functional specialization and decreasing sectoral specialization is confirmed by empirical study of specialization of US states over a 20-30 year period.