New research insights on firms, trade, and development

Firms affect poverty and economic growth in low- and middle-income countries (LMICs) not only through their output but also because of their role as employers. For example, manufacturing firms affect the economy both through the products they produce and because they provide wages and other services for employees. 

Working for a firm can often generate higher incomes than achievable through a microenterprise or agricultural production. This role of firms as employers is important for LMIC households with limited resources, since their time may be the primary – if not only – asset to generate earnings, especially for non-agricultural households. And moving workers from agriculture to firms can have broad economy-wide benefits as the key to the structural transformation

Understanding labour market outcomes, including firms’ role as employers and the determinants of labour supply and productivity, has important implications for both the livelihoods of people with low income and economic growth. These topics will be at the centre of discussions at the  Firms, Trade, and Development Conference 2024, co-hosted by the Economic Growth Center and the International Growth Centre at the Yale University on 24 and 25 October, with a live Zoom simulcast. 

Emerging insights in the current firms, trade, and development research 

Looking at some of the exciting research that will be presented at the conference, four key insights emerged. In my discussion below, I provide links to the presenters’ websites and to working paper versions that are available online. Some drafts may be preliminary and change before conference presentations.

1. Firms can have large effects on worker productivity through shaping the work environment and incentives. 

Firm choices influence the work environment and the incentives that workers face, with important implications for worker productivity. 

Firms can make important investments that shape the work environment to improve worker productivity. Maulik Jagnani (Tufts University) and co-authors study how labour productivity responds to air purifiers in garment manufacturing firms in Dhaka, Bangladesh, where indoor air pollution is far beyond the World Health Organization’s safe limit. By comparing output from workers at firms that randomly receive air purifiers with those that do not, they show that improvements to air quality from air purifiers improve productivity by 10%. 

Deivy Houeix (MIT) emphasises how digital technologies bring direct benefits but also increase the ability of firms to monitor their employees’ efforts. Studying the taxi industry in Senegal, he finds that randomised access to a digital payment technology, which allows taxi owners to observe driver behaviour more easily, increases worker effort by almost 30% and changes contracts. But many of the poorest drivers choose not to adopt unless visibility is curtailed. 

2. Firms’ and workers’ individually optimal decisions can have aggregate effects that imply markets produce suboptimal outcomes.

Decisions by individual firms and workers that affect labour supply and productivity can affect others in meaningful ways through externalities. In the absence of policy interventions, when firms and workers do not incorporate the effects of their actions on others in decision-making, markets can fail to maximise efficiency.

Yogita Shamdasani (National University of Singapore) and co-authors illustrate how habits shape labour supply. Specifically, individuals who work more – due to random exposure to work bonuses – increase future labour supply. Similarly, workers who encounter circumstances that lead them to exit the labour force, such as funerals, reduce subsequent labour supply. It follows that individual firms do not capture the full benefit of providing employment opportunities. Current work opportunities increase future labour supply, which also benefits other firms, generating a positive labour-supply externality. Therefore, levels of work and wages may be lower than what would maximise long-run economic gains.

Studying food delivery workers in urban China, Yulu Tang (Dartmouth) quantifies two important externalities of geographic clustering. First, productive knowledge transfers are concentrated among workers from identical hometowns – a positive externality. Second, congestion from workers simultaneously increases labour supply in response to adverse events in hometowns such as floods, which reduces wages during this time of need – a negative externality. As workers do not individually consider these positive and negative externalities when choosing their work location, the clustering of workers can be suboptimal.

3. The business conditions and environment within which firms operate matter for labour productivity and supply.

The business environment can also directly affect workers and labour productivity. Due to challenges in identifying far-away trading partners, limited information about other firms, and other factors, many LMIC firms only sell locally. Even though firms might sell only locally, they often need a full “unit” of labour (or capital) to make even one sale, as highlighted by Tilman Graff (Harvard University) and co-authors. For example, a retail shop always needs at least one worker, even for stores that are mostly empty. This “integer problem” (i.e., that some decisions require a full next unit) can leave resources, including labour, underutilised, adversely impacting productivity.

Wei Qian (Haverford College) and co-authors illustrate that geographic integration of economic activity affects labour supply and productivity. On the one hand, improved integration enables firms to expand and access new demand. However, increased competition can close firms, restraining movement from low-productivity subsistence agriculture. On the other hand, increased access to agricultural products for consumption from elsewhere can release labour from the agricultural sector, increasing non-agricultural employment. Empirical analysis of major highway constructions in India and China shows that both forces matter. In India, where the agricultural share was relatively low, the first channel dominates, while the opposite occurs in China.

4. Worker’s assets and knowledge can play an important role in determining their productivity within the firm, even through unexpected channels.

Workers’ characteristics, both in terms of their assets and knowledge (i.e., both physical and human capital), can affect the performance of the firms they work for. In LMICs, where firms may face challenges in acquiring their own assets or training workers, these worker features might be especially important.

Jiayue Zhang (Brown University) and coauthors highlight that many LMIC workers bring productive machinery (“capital”) to work: some garment workers bring sewing machines and some carpentry workers bring wood planers. When some workers randomly receive capital due to a cash transfer, worker productivity increases. Suggestive evidence highlights that productivity improvements may reflect firms responding to the capital by increasing revenue-sharing. Then, workers increase effort because of improved alignment of incentives. It follows that firms could benefit by helping workers access capital, such as dedicating some of a worker’s compensation to it. And, because of overall productivity gains, the firm and worker could both gain if the firm subsidises access to capital.

Finally, how does low-skilled immigration from a high-income country to a LMIC affect firm productivity? Daniel Osuna-Gomez and a co-author study the effects of US deportees to Mexico on firm behaviour. They find that firms in areas with more deportees benefit relative to comparison firms: they report higher survival rates and average revenue. Firm benefits increase in the US tenure of the deportees point towards the benefits of knowledge and financial cross-border transfers from workers to firms.

What are the policy implications of this research? 

I propose four key takeaways:

1. Policymakers may consider targeting barriers that limit firms’ adoption of technology and practices that improve productivity. Such barriers may include limited knowledge of the existence of certain technologies and practices, or their benefits, as well as financial or other implementation barriers that firms face.
2. Policy design needs to mitigate externalities. For example, when habit formation has large labour supply effects, wage and work subsidies could counteract suboptimally low levels of work. Similarly, providing access to insurance can help workers respond to negative events without causing congestion externalities.
3. While geographic integration can spur economic growth through many channels, it also redistributes productive resources and opportunities. Policies can help those adversely affected by integrating access to resources to offset losses.
4. As firm and worker productivity can depend upon many features of the employer-employee relationship and match, policymakers need to evaluate how to best support firms and workers to access assets and knowledge that improve productivity.
    

This article was published in collaboration with the Yale Economic Growth Center to mark the upcoming Firms, Trade, and Development conference.