Understanding low productivity in developing countries: Evidence from the airline industry

Economists have long been interested in documenting and explaining the gap in productivity between developing and developed countries. Understanding the determinants of low productivity in the developing world is important not just for economists. This knowledge is vital for policymakers in government, our primary stakeholders, in order to guide the design of industrial policies and institutional reforms (and to allocate scarce government resources) to close this productivity gap and spur economic growth.

The goal of this project is twofold: The first goal is to produce a new set of comparable productivity measures for many firms in many countries. Early work on this subject relied on national accounts data to calculate productivity measures across countries and while these studies generated important findings, they necessarily required strong assumptions on data comparability and production structures.

This project circumvents these data comparability challenges by focusing on one particular industry that is present in virtually every country in the world and for which there exists rich and comparable data: the airline industry. Moreover, an additional and important benefit of analysing the airline industry is that multiple airlines from different countries often fly the exact same routes. This allows us to construct direct, clean and comparable measures of performance based on flight delays, accidents, capacity utilisation, and revenues. The airline industry is unique in having governing bodies that report standardised and comparable data across countries and time. In addition, we can assess the productivity of airports and routes by comparing performance of the same airline across multiple airports and across multiple routes.

Using the productivity measures from thousands of airlines and airports across virtually all countries in the world over time, the second goal is to explore how the levels and distribution of productivity co-vary with both country and firm characteristics. For example, literature on growth and institutions have hypothesised that country-level characteristics, such as property rights, rule of law, and human capital, drive cross-country differences in productivity (e.g., Hall and Jones 1999).

We also intend to exploit various policy and industry changes during the sample periods (such as the signing open skies agreements, foreign takeovers and privatisations) to ascertain the direction of causality for some of these associations. Finally, we plan to explore how institutional characteristics interact with firm-level characteristics.