Should Capital Flow from Rich to Poor Countries?

Abstract

Are human and physical capital stocks allocated efficiently across countries? To answer this question, we need to differentiate misallocation from factor intensity differences. We use newly available estimates on factor shares from Monge-Naranjo, Santaeulàlia-Llopis, and Sánchez (2019) to correctly measure the factor shares of physical and human capital for a large number of countries and periods. We find that the global efficiency losses of the misallocation of human capital are much more substantial than those of physical capital, amounting to 40 percent of the world's output. Moreover, contrary to the findings of Monge-Naranjo, Santaeulàlia-Llopis, and Sánchez (2019) for physical capital, the global misallocation of human capital does not seem to be subsiding. We argue that the proper measure of global misallocation requires considering the potential gains of reallocating both physical and human capital. In this case, the implied efficiency loses from misallocation are up to 60 percent of global output. Attaining those gains, contrary to the prominent Lucas paradox (Lucas, 1990), would often require physical capital to flow from poor to rich countries.

Introduction

The large dispersion in real wages across countries suggests a potentially huge global misallocation of human capital. Thus, reallocating human capital could substantially increase global output and drastically change the world income distribution. To be sure, reallocating humans across countries is a much more complex endeavor than reallocating physical capital. Migrant workers, and not machines, leave behind friends, families, and other attachments and may face cultural and anti-immigrant resistance. Moreover, the impact—real or perceived—of foreign workers on the local population has been used as a political banner that has no counterpart with the impact of capital inflows. Yet, despite all those frictions and barriers, workers and their human capital have been continuously reallocated across countries, oftentimes in great measure. As of today, in the United States and in many other countries, such a reallocation is evident not only in high-human-capital-intensive institutions, such as universities, hospitals, and research institutions, but also much more generally in stores, restaurants, and farms, all of which often agglomerate workers from all over the world.

In this article, we assess the potential global efficiency gains and distributional impacts of reallocating human capital across countries. To this end, we face a number of challenges. First, we need to take a stand on which factors are fixed in each country and which factors can be reallocated—if any—including human capital. Second, we need to control for factor intensity differences across countries to avoid confusing them with distortions. Third, we need to measure or infer the marginal valuation of human capital across countries and incorporate some of the distributional constraints that countries may impose for the entry of workers from abroad. We use the recent work by Monge-Naranjo, Santaeulàlia-Llopis, and Sánchez (2019) that provides exactly the data required to address these three issues for a sample of 76 countries and for the years 1970 to 2005. First, aside from pure total factor productivity (TFP), natural resources are ultimately the only fixed inputs of production in each country. Using the measures in Monge-Naranjo, Santaeulàlia-Llopis, and Sánchez (2019), we assess the curvature of the production function of the different countries with respect to all the mobile factors, that is, human and physical capital, and evaluate the gains of reallocating human capital only or human and physical capital simultaneously. Second, we use the measures in Monge-Naranjo, Santaeulàlia-Llopis, and Sánchez (2019) to control factor intensity differences across countries, which they show are not sensitive to policy distortions. Third, we circumvent the lack of direct and reliable measurements of the relative value of human capital across countries and periods, using the model to generate two extreme and opposite bounds for the observed costs of labor across countries.

Our basic efficiency benchmark consists of equating the marginal returns to human capital across countries. Doing so points to large misallocation of human capital during the sample period, in the range of 40 to 50 percent of global output, with an upward trend over time. Our findings resemble those in Klein and Ventura (2009) and Kennan (2013), using different models, countries, and data. This basic benchmark abstracts from the barriers to reallocating human capital (workers) across countries, which can be very stringent. Some of the barriers are natural, such as the emotional cost of reallocating human beings across countries with different languages, cultures, and values. But other barriers are the result of policies and legislation, mainly in the more developed countries. Such barriers are surely motivated to prevent a reduction in the wages of some of the domestic workers. In fact, the large implied global output gains from the basic benchmark come at the cost of drastic reductions in the wage rate (per unit of human capital) in developed countries.

To appraise the potential gains in global output without the negative impact on the domestic workers of developed countries, we construct policy counterfactuals that are constrained so that the real wages of workers must be kept constant (at the implied levels from the data). By design, if workers were the only factor that could be reallocated across countries, no reallocation would take place and global gains would be zero. However, if both human and physical capital could be reallocated, even under such a conservative exercise, the global gains would be substantially higher than reallocating physical capital alone, around 8 percent to 9 percent of global output in the 1970s and up to 6 percent by the 2000s. Interestingly, the reallocation is largely from the richer and poorer countries (first and fourth income quartiles) toward the middle ones (second and third income quartiles.)

Overall, a proper assessment of global misallocation considers both human and physical capital. The complementarity between these two factors plays a role, as they must be directed toward the countries with higher fixed productivity, either because of TFP or natural resources. Observed allocations deviate from such an alignment. More interestingly, if human and physical capital can be reallocated jointly to equalize their marginal returns across countries, the direction of the physical capital flows can be reverted relative to the case when physical capital is the only mobile factor. In fact, the premise that capital should flow from rich to poor countries is unwarranted: When both factors are reallocated, capital and labor would flow from some of the poor and middle-income countries toward some of the richer countries. This simple yet often ignored point could be one of the keys to understanding the consequences of alternative integration schemes with or without labor mobility for countries and regions with different productivities and fixed endowments (e.g., the United States and Puerto Rico and the European Union on one side, with NAFTA on the other).