Evidence from Better Factories Cambodia
Low wages, long hours, high temperatures, excessive noise, poor air quality, unsanitary conditions, and abuse (both verbal and physical) in developing country manufacturing establishments are often cited as evidence that “sweatshops” characterize production in relatively poor countries. Harsh working conditions in apparel factories are at the center of a large and growing debate about globalization and labor standards (Elliott and Freeman 2003). Several organizations have responded to rising public concern by pressuring governments and employers to improve working conditions. Public exposure, such as anti-sweatshop agitation, seems to have improved working conditions in global supply chains. Harrison and Scorse (2010), analyzing Indonesian manufacturing census data from 1990s, find that workers in the apparel, textile and footwear global supply chains were underpaid relative to workers supplying the domestic sector prior to the anti-sweatshop campaign of the early 1990s, but, by 1997 were earning more than the comparison group.
One concern about anti-sweatshop activity is that it imposes constraints on factories that, especially in very competitive environments, may make survival more difficult. Many improvements require costly capital investments (such as air conditioning, plumbing, or safety equipment). Complying with minimum wage laws and providing additional compensation (such as paid leave and overtime) also can increase factory costs. These arguments are trivial to illustrate using even the most basic economic theory: if firms are operating efficiently in competitive markets, increases in costs (holding all else constant) will necessarily cause marginal firms to exit.1
Factory closings are a considerable concern in developed and developing countries. As a result, there is a sizable literature that seeks to uncover the variables linked to factory survival. Early papers focused on the United States and other developing countries (Bernard and Jensen 2007, Disney et al. 2003, Doms et al. 1995, Baggs 2005, and Greenaway et al. 2008). These papers illustrate the importance of technology, capital intensity, age, and size in survival rates. Recent papers, such as Harris and Li (2010) find a positive relationship between exporting in particular or exposure to foreign markets generally and survival. Advances in data collection and availability has extended this literature to developing countries, including Ghana, Kenya, and Tanzania (Soderbom et al. 2006), Ethiopia (Shiferaw 2009), Indonesia (Behrman and Deolalikar 1989), and Malaysia (Nor et al. 2007). In general these papers also find positive effects on survival of technology, size, and links to foreign markets.
Few, if any, of these papers identify underlying driving factors of technology that may be improving survival. Fewer still examine the link between changes in human resource practices in general, much less improvements in working conditions, with changes in survival probabilities.
The goal of this paper is to analyze changes in working conditions in Cambodian apparel exporting factories to see which, if any, are statistically related to the probability of factory closure. Our very preliminary results suggest that most dimensions of working conditions are not statistically related to closure: only 3-4 of the 31 groups show evidence of a statistically significant relationship (depending on the estimation method and specification). Those that are significant (or nearly significant) tend to be those that are most directly related to worker effort – weekly hours, weekly rest, and payment of wages. We find very limited, if any, evidence that improvements in any of the areas increase the probability of closure in a statistically or economically important way. These preliminary results seem inconsistent with the argument that improving conditions puts unbearable cost pressure on factories.