Time to Rebalance Competition Policy for Workers?

Economic theory suggests that in a competitive labour market, the wage paid to the worker should be equal to their marginal product, however, evidence shows that firms can and often do pay wages well below the marginal product – resulting in a markdown in wages.

This markdown is prevalent not only in high-income countries such as the United States (US), where the average manufacturing worker earns only 65 cents on the marginal dollar generated, but also in large developing countries, such as China and India, where workers receive wages of 16% and 18% lower than the competitive level, respectively.

Such large markdowns reflect excessive employer-labour market power, which is beneficial to firms, but detrimental to workers and the economy as a whole.

By artificially lowering wages, employer-labour market power reduces the economy’s employment and output.

That is because labour supply is typically upward sloping, hence a lower wage is associated with lower equilibrium employment.

In this context, the additional employment induced by a wage increase would generate an increase in output still above the wage, thus enhancing overall welfare.

The impacts are sizable, with evidence from the US suggesting welfare losses from labour market power of around 8% of gross domestic product (GDP) and output losses equivalent to a fifth of GDP.

Moreover, these figures do not account for dynamic effects on growth, such as reduced human capital investments due to lower wages.

In addition, by depressing employment and wages, labour market power also reduces the labour share of national output, a key measure of inequality that has been declining in most of the world, including developing countries.

MARKET REGULATION

To reduce the negative impact of employer-labour market power, it is crucial to understand its causes.

Previous research has identified employer concentration as the key contributor. In concentrated labour markets workers have fewer employment options, which increases the relative bargaining power of employers.

Empirical evidence supports this hypothesis, showing lower wages in more concentrated labour markets in the US and Peru.

As a result, some experts have suggested using antitrust approaches to regulate labour markets.

However, the evidence of a causal relation between market concentration and labour market power remains unclear.

In fact, the correlation between concentration and markdowns at the aggregate market level has also been questioned in the US.
To help fill this evidence gap, a recent paper focused on assessing the impacts of changes in barriers to entry in product markets on labour market power in Indonesian manufacturing.

Our estimates suggest that Indonesian plants exert on average some degree of labour market power.

The study leveraged Indonesia’s negative investment list (NIL) to measure changes in investment conditions across narrowly defined product markets.

The granular data collected for successive iterations of the NIL in the late 2000s to early 2010s allowed us to show that when at least one investment restriction is introduced in a product market, firm entry in that market drops by around a quarter.

As a result of the lower entry, the wage markdown increases by the same proportion.

At the same time wages go down and price markups go up – a sign of declining competition.

The main findings are statistically significant at conventional levels and are robust to a wide array of tests addressing various estimation concerns, including the endogeneity of investment restrictions and the definition of labour markets.

COMPETITION POLICY

The evidence suggests that product market regulation has a significant impact on labour market power through its impact on firm entry, and hence on labour market competition.

The existence of this negative labour market externality of product market regulation supports the calls to evaluate any change in industry concentration, due, for instance, to mergers or investment restrictions, not only against its impact on product markets, but also on labour markets.

This has two key policy implications.

First, it calls on competition authorities to extend antitrust approaches from product to labour market assessments – something they have been largely neglecting so far. This should not require significant technical efforts.

Naidu, Posner and Weyl propose relatively straightforward adaptations to labour markets of three standard approaches to product market merger analysis.

For example, adapting standard approaches to product market merger analysis, such as the hypothetical monopsonist test and the Herfindahl-Hirschman index of labour market concentration, can be useful for assessing labour market power.

Second, the findings in the paper highlight the key role governments can play in reducing firms’ market power in labour markets by regulating product markets, for instance, reducing investment restrictions.

Those are standard policy tools that go beyond antitrust regulation and competition authorities, but which can be as effective to curb excessive labour market power.

• Massimiliano Calì is the senior country economist for the World Bank in Tunisia, and Giorgio Presidente is a postdoctoral researcher at Oxford University.