The commodity price boom and regional workers in Chile: a natural resources blessing? Andrea Pellandra⇤ March 11, 2015 Abstract This paper uses variation in Chilean regions’ exposure to a large exogenous price shock to examine the effects of the growth of a geographically concentrated natural resources-intensive export sector on local wages and employment between 2003 and 2011. I develop a specific- factors model of local labor markets with two types of workers (skilled and unskilled) that are mobile across sectors within a region but imperfectly mobile across regions that predicts that a price shock in an industry that intensively employs unskilled labor will reduce local wage premia proportionally more in regions where that industry represents a higher share of total employment compared to other regions. Empirical results show that a region exposed to a 10% increase in average prices (or a 10 dollars per worker exogenous increase in exports) experienced a2.4%increaseinaverageunskilledworkers’wagesrelativelytootherregions,andthatsuch gains contributed to a reduction in regional wage premia. I also find substantial employment reallocation from unaffected sectors to the sector that benefited from the price increase within regions, but much less evidence of mobility across regions. Finally, there is clear evidence of adeclineinpovertyratesandincomeinequalityinregionsmostexposedtothepriceshock compared to other regions. JEL Classification: F14, F16, O15 1 Introduction Even though large endowments of natural resources can unarguably provide vast opportunities for economies to grow and develop, the impact of natural resources wealth on economic development is still the subject of ample debate among economists. The traditional question researchers have tried to answer is why some resources-rich economies, such as Australia, Canada, and Norway, have performed well, while others, such as the Congo, Nigeria, and Venezuela, have not been successful; ⇤Author’s affiliations: Heinz College, Carnegie Mellon University, 4800 Forbes Avenue, Pittsburgh PA 15213. Electronic-mail: [email protected]. Economic Commission for Latin America and the Caribbean, Dag Ham- marskjold 3477, Santiago, Chile. Electronic-mail: [email protected]. 1 The Commodity Price Boom and Regional Workers in Chile Andrea Pellandra however, in recent years, a number of studies have also started to look at the local economic effects of natural resources booms within the same country. Indeed, each country is made up of regions that vastly differ in their performance and growth rate, and local labor markets are characterized by huge disparities in terms of factors’ productivity, firms’ innovation and workers’ employment and wages1. Understanding the specific labor market conditions of these regions and responding with policies that incorporate local solutions is therefore a key element in the debate over national economic development. The recent debate on the local effects of natural resources booms has mainly focused on trying to determine whether local shocks that increase oil, gas or mining production benefit producing regions by increasing local employment and wages, or if they rather crowd out other sectors of the economy such as manufacturing by increasing factor prices. The issue is not trivial, as it is closely related to that of the localization of economic activity and its effects on regional prosperity and workers’ welfare which has intrigued urban economists for decades. The main explanation for the reason why firms tend to cluster in particular areas has been found in the existence of agglomeration economies that can generate indirect local “spillover” effects: these local spillovers can be a sizable benefit of firms’ location decisions, and if manufacturing generates positive spillovers that natural resources firms do not, local workers may end up worse off. Unfortunately, quantifying the effects of a shock to a local labor market can be problematic as firms’ decisions are endogenous and therefore the counterfactual is missing. In this paper, I exploit the variation in the industry mix of employment across Chilean regions to identify the effect of differential regional exposure to a large exogenous commodity price shock on wages, employment, poverty and inequality at the local level. My results show that average wages increased the most in local labor markets where employment was more concentrated in industries which presented larger price increases, with regions facing a 10% increase in regional prices experiencing a 1.5% wage increase relatively to other regions. Such wage gains can also be expressed in terms of the increase in exports, and I show that in regions that experienced a 10% exogenous increase in exports due to the price effect, workers’ wages increased by 1.6%. However, as a the natural resources-intensive sector that experienced the price shock absorbed a higher proportion of 1As pointed out in a recent OECD report (OECD (2013)), “each job is local”: in 10 OECD countries, more than 40% of the increase in unemployment over the past five years was concentrated in just one region, with regional disparities in youth unemployment growing even wider. 2 The Commodity Price Boom and Regional Workers in Chile Andrea Pellandra the regional unskilled workers than skilled workers, most of the benefits in terms of higher local wages accrued to less educated workers, while the increase in regional prices and exports had a much more modest effect on the wages of college educated workers. Therefore, such gains contributed to a higher proportional reduction in wage premia in affected regions compared to other regions. I also find substantial employment reallocation from unaffected sectors to the sector that benefited from the price increase within regions, and that even though there is a much lower evidence of mobility across regions, due to the differential migration patterns across skill groups the overall decrease in wage inequality was lower than it would have been in absence of interregional mobility. Finally, there is clear evidence of a decline in poverty rates and income inequality in regions most exposed to the price shock compared to other regions. The Chilean economy is characterized by wide regional disparities, due to its high dependence on few primary sectors - especially copper - concentrated in limited regions. Territorial inequalities among regions in Chile are the highest of all OECD countries, with the richest region generating over eight times the regional output per capita of the poorest one2. Figure 1 shows the GDP per capita in 2005 PPP constant thousands of dollars of Chilean macro-regions in 2003 and 2011, whilst as a reference, Figure 2 compares 2011 GDPs per capita of some regions with those of other world countries. As can be clearly seen, the per capita GDP of the poorest Chilean region, Araucania, is lower than that of Egypt, while the one of the richest region, Antofagasta, is higher than Switzerland. The GDP of the Santiago region, by far the largest in terms of population, approximates the Chilean average and is comparable to that of Croatia. The reason behind such disparities is the high dependence of Chile on natural resources: output is mainly generated in the areas where such resources are located. In fact, the three regions with the highest GDP per capita, Antofagasta, Tarapaca, and Atacama, are also the regions with the largest deposits of copper, by far Chile’s main produced and exported commodity. Therefore, due to the geographical variation in copper production, and the large, exogenous increase in world prices experienced by copper in the past decade, Chile arguably provides an excellent setting to study how a natural resources windfall can affect local labor markets and regional economic development. The rest of the paper is organized as follows. In section 2, I begin by reviewing the recent 2These are huge differences: by comparison, the GDP per capita of Delaware, the richest state in the United States - a country also characterized by strong territorial inequalities - is just over twice that of Mississippi, the poorest state. 3 The Commodity Price Boom and Regional Workers in Chile Andrea Pellandra empirical literature, starting from the popular and ubiquitous one on the “natural resource course” that has nevertheless proved to be quite elusive to sound empirical evidence. In section 3, I introduce a short description of the importance of the copper sector in Chile, and present evidence of the role played by Chinese demand in stimulating the stunning boom in world prices experienced by the commodity in the first decade of this century. Section 4 introduces the theoretical model, which adapts Jones and Marjit (2003)’s 4x3 specific factors model to allow for a spatial dimension, to include a non traded sector and to consider two types of mobile labor, and which I use to guide my empirical analysis. In section 5, I describe the data, and introduce some descriptive statistics to show the evolution of the economic structure and of average wages across Chilean regions. In section 6.1, I use average regional price changes to estimate the effect of the regional price shock on normalized regional wage premia, calculated from Mincerian wage regressions, while section 6.2 uses a similar approach to estimate the impact of an increase in regional exports on local wages; section 6.3 completes the analysis examining the migration patterns of workers of different educational levels. In section 7, I extend the analysis to calculate regional wage gaps