Why the skill composition of trade unions matters for wage inequality


by Kyung Joon Han and Eric Graig Castater

There has been a longstanding consensus in the labor politics literature that countries and time periods with higher union density (the proportion of workers that belong to a union), union coverage (the proportion of workers covered by a union bargained contract, regardless of whether they are union members), and/or wage bargaining level (the level at which unions and employers bargain over wages, ranging from the firm to the national level, with the latter signifying a “higher” level of wage bargaining) are associated with lower levels of wage inequality.

Indeed, much of this scholarship has found that such “union presence” variables help explain a substantial portion of the cross-national and over-time variation in wage inequality in the wealthy democracies. However, recent evidence indicates that such a union effect on wage inequality disappeared during the 1990s due to the decentralization of bargaining structure as well as the decline of union density and union coverage.

In short, unions are only able to reduce wage inequality when they have substantial organizational strength.

In an article recently published in Research in Social Stratification and Mobility, we examine whether unions still have an effect on wage inequality, but do so by focusing on what we term the capital and skill mechanisms, the primary means by which unions reduce wage inequality.

The capital mechanism refers to the ability of unions to extract wage concessions from their employers. This may be done, for example, by workers jointly withholding their services until particular demands are met regarding compensation.

The skill mechanism refers to the ability of less skilled union members to extract wage concessions from more skilled union members. This is done when union members bargain among themselves before (and likely while) they bargain with employers. Such bargaining is generally how unions set their goals (e.g., on wage and employment levels) for their negotiations with employers.

Thus, union goals are shaped by conflict and compromise between (union) workers with unlike preferences.

Despite the (implicit) recognition of both of these mechanisms in the labor politics literature, the union variables scholars have employed to determine union effects on wage inequality (union density, union coverage, and the level of wage bargaining) are unable to measure the skill mechanism, as they do not consider what types of workers are actually unionized.

In order to test whether unions are still able to activate the skill mechanism, we constructed our own data of the proportion of union members that are unskilled manual workers using survey data.

Based on the assumptions that union member preferences are substantially determined by their skill level (with less skilled union members desiring less wage inequality than more skilled union members) and union decision-making is approximately democratic, we find that greater numbers of unskilled union members relative to more skilled union members increases unions’ ability to activate the skill mechanism and thus reduce wage inequality between the years 1990 and 2009: as the proportion of union members that are unskilled manual workers rises, wage inequality falls, regardless of the level of union density or the level of union coverage.

However, the share of such workers has no effect on wage inequality when wage bargaining institutions are decentralized (i.e., occur at a “lower” level), likely because unskilled manual workers are unable to extract wage gains from their more skilled and higher paid union brethren in such an institutional context.

Why is the level of wage bargaining the only union presence variable to have an impact on the unskilled manual union member effect on wage inequality? It is likely because the centralization of wage bargaining variable is the only union presence variable that explicitly and exclusively measures an organizational characteristic of unions.

When unskilled union members are in the same union organization as more skilled union members (i.e., a more centralized wage bargaining structure), they are able to extract wage concessions from them. By contrast, when they are in a more fragmented union structure, as a lower level of wage bargaining implies, it is likelier that each union is more homogenous in regards to occupation and skill, and thus the skill mechanism cannot be activated.

Yet even if union organizations represent a more heterogeneous group of workers in such a setting, the plethora of exit options for more skilled union workers in a decentralized context likely precludes the ability of less skilled union workers to substantially activate the skill mechanism. Furthermore, the unequal distribution of resources among different union organizations may even allow unions with a larger proportion of higher skilled (and thus better paid) union workers to achieve greater gains through collective action (whether toward employers or politicians) than their less skilled (and thus lesser paid) union brethren.

Our findings have at least four important implications. First, unions do have an effect on wage inequality even in the post-1989 period, but this effect is driven not just by the redistribution of earnings from employers to workers, but also from some union workers to others. Second, we need to consider the preferences and behavior of different types of union members if we are to properly account for the impact of unions on political and socio-economic outcomes. Third, even though politicians may be more responsive to the policy preferences of the wealthy, poorer individuals can achieve relative economic gains if properly organized.

Finally, if activists and politicians want to blunt the trend toward greater wage inequality, they should focus their efforts on organizing less skilled workers, particularly those whose functions serve as strong complements to more skilled workers. This would ensure that unskilled workers hold substantial bargaining leverage and would mean more emphasis on organizing in industries such as health care, in which highly skilled workers (e.g., nurses and doctors) substantially rely on the hard work and assistance of less skilled workers (e.g., nurses’ aides).

Kyung Joon Han is an assistant professor in the Department of Political Science at the University of Tennessee. Eric Graig Castater is a lecturer in the Department of Political Science at the University of Tennessee.

This article summarizes findings from “They May Not Have the Skills, but They Have the Desire: Why the Skill Composition of Trade Unions Matters for Wage Inequality” in Research in Social Stratification and Mobility.

Image: The All-Nite Images via Flickr (CC BY-SA 2.0)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: