Aggregation in Gravity-Based Estimation: Theory, Evidence and Policy Implications

International trade forms a vital part of modern economies. Likewise, many aspects of government policy are directed towards influencing international trade. For instance, governments may join trade agreements, enter currency unions or lower tariff barriers. It is therefore crucial to understand and predict the effects of such policies on international trade, particularly at a time when the UK government is looking to embark on the (re)negotiation of multiple trade deals. The so-called gravity equation is the most widespread and influential tool to estimate these effects in the data.

One difficulty in understanding the behaviour of trade in response to government policy is that trade data are typically aggregate statistics. Aggregation occurs with respect to two major dimensions. First, aggregation can apply over different industries, for instance when exports of cars and furniture are combined. Even within narrowly defined industries, trade data as reported in standard statistical data sets typically aggregate over different firms and their transactions. Second, aggregation can refer to space. For instance, exports from Britain are aggregates of exports from individual regions.

The aim of our project is to shed light on how aggregation affects estimates in gravity estimation, both from a theoretical and an empirical viewpoint. When trade policies change, they typically affect individual economic agents at the micro level. For example, when a tariff is introduced, it affects the behaviour of firms and consumers. Gravity estimation, however, is based on aggregate data. To what extent can we then use gravity estimates to identify relevant parameters that operate at the micro level? Identifying these structural parameters is of fundamental importance, for instance for policymakers trying to understand the effects of trade policy.

The response to a trade policy change such as a new trade agreement might differ at the micro level compared to what aggregate gravity estimates typically suggest, either because firms operate in different industries with different characteristics, or because firms are based in different locations (say, closer to the border or further away) and therefore react differently.

Apart from contributing to the academic literature, our research aims at improving the accuracy and predictive power of empirical trade models for practical purposes. It will allow policymakers to better forecast the effects of proposed policy changes. This is particularly important for the UK government in the wake of the Brexit referendum where it is likely that many existing trade agreements have to be renegotiated or new ones devised. Our research can also help to understand the exporting behaviour of firms, and ultimately help to identify policies that show more promise to increase exports.

Apart from policymakers in government, we also see our research as beneficial to international organisations such as the World Trade Organization and the OECD. Those organisations are concerned with the effects of tariffs and non-tariff barriers on international trade. We have strong existing links to officials from these organisations and will use them to disseminate and discuss our research.

Apart from academic beneficiaries, our research will benefit policymakers from a range of institutions. Our research will be directly relevant to any government department or international organisation involved in the design and evaluation of trade policy. In the UK in particular, this includes HM Treasury, BEIS, UK Trade & Investment and the new Department for International Trade. In the wake of the Brexit referendum, for the first time in over 40 years the UK will have to build up its capacity to negotiate trade deals, and this process will have to rely on expert advice and up-to-date techniques.

Other potential users are other governments, the European Commission, the Organisation for Economic Cooperation and Development (OECD), the World Trade Organisation (WTO), the International Monetary Fund (IMF) and the World Bank. These institutions rely on the gravity equation to evaluate trade policy arrangements, for instance to predict the effect of new trade agreements, and we have professional contacts with them.

We also expect impact for industry and consumer organisations such as the British-American Business Council (BABC) and associated chapters since they also routinely evaluate trade policy.