When unions succeed in raising worker wages, who ultimately bears the cost? The question has animated labor economics for decades and has gained fresh policy salience as U.S. union density has fallen to roughly 10 percent of workers—about 6 percent in the private sector—while recent organizing campaigns at companies such as Amazon, Starbucks, and the major automakers have renewed attention to collective bargaining's effects on firms, workers, and consumers. A forthcoming study in the Quarterly Journal of Economics by Samuel Dodini, Anna Stansbury, and Alexander Willén provides a comprehensive assessment of firm-level responses to increased unionization, drawing on data from Norway, where union density remains among the highest in the OECD at roughly 50 percent and labor-market institutions provide a distinctive setting for identifying causal effects.

The researchers exploit changes in the tax deductibility of union dues in Norway as a quasi-exogenous source of variation in firm-level union density. Their findings reveal that the cost of higher union density is shared among several stakeholders, but the distribution depends heavily on market structure.

In the average private-sector firm, higher union density raises labor costs and leads firms to contract both employment and production, which lowers profits without increasing the labor share. The incidence is shared across groups: consumers bear part of the cost through higher prices, shareholders absorb losses through lower profits, and a portion is offset by productivity improvements. The total wage bill falls, with losses concentrated among less-attached "outsider" workers—a result that connects to longstanding debates about how unions may advantage incumbent "insiders" at the expense of marginal or prospective entrants to the labor market.

Firm responses vary systematically by the degree of market competition. In the manufacturing sector, where firms operate in less competitive product and labor markets, the response reverses. The average manufacturing firm expands employment and production, reduces labor markdowns, and does not experience profit declines. Instead, higher labor costs are largely passed on to consumers through higher prices, with the remainder offset by productivity gains. In this setting, workers benefit as both wages and employment rise. The finding that unions compress labor markdowns—the wedge between wages and marginal revenue product—adds empirical weight to a growing literature on employer monopsony power, which has documented that many firms possess meaningful wage-setting discretion.

These patterns suggest that unions can offset employer monopsony power, and that firm responses—and therefore who ultimately bears the cost—depend importantly on market structure. Overall, the study finds that unionization in Norway primarily redistributes costs from consumers rather than shareholders, and produces effects that differ sharply across firms, including a reallocation of economic activity toward larger and more productive firms.

The authors rationalize these patterns using a partial-equilibrium model of union bargaining that incorporates both product- and labor-market power. A key question for future research will be the extent to which these mechanisms, identified in Norway's high-coverage, centralized-bargaining environment, translate to economies where bargaining is more decentralized and union density is substantially lower.

The paper is forthcoming in the Quarterly Journal of Economics (abstract).