Much of the best sociology of pay-setting from the last twenty years has documented the declining importance of non-market constraints on pay.
The collapse of labor unions means fewer and fewer employers are bound by restrictive collective bargaining agreements. Similar outcomes come from the rise of shareholder value; restructuring and opening up of internal labor markets; erosion of fixed, bureaucratic pay schedules; outsourcing and contracting…The list goes on.
There’s a ton of great research on these important changes. But they can mostly be summarized as changes that eliminate protections from market competition that workers previously enjoyed.
And the consequences of the increased exposure of workers to the competitive labor market has been pay stagnation, rising wage inequality and zero progress on closing the black/white income gap. Important and bleak stuff; echoes of Polanyi throughout this subfield.
Now to be clear, most of my own work (like on labor market institutions in the 1970s, on the rise of buyer power in supply chains and the rise of merit pay) is totally consistent with this story. So, I think there is a lot that this work gets right.
But I want to focus on a different and totally parochial consequence of these changes. Economic changes seem to be paring back the traditional jurisdiction of sociologists. If non-market constraints are disappearing, should we just focus on understanding labor supply and demand? Should we become labor economists?
In a recent article in Administrative Science Quarterly, I put forward an alternative. The starting point for the article was a hypothesis that the core insight of the sociological study of pay setting is not exactly about structural constraints on market forces.
It’s actually more general than that: non-market interdependencies among workers can affect pay, even if these interdependencies aren’t reducible to bureaucratic employment policies, collective bargaining agreements or other structural constrains. This puts the issue of externalities among workers (and between workers and employers) at the center of pay-setting.
Ok, that is very abstract. What do I mean by interdependencies? An obvious is the social or relational interdependency among people trying to collaborate on completing some work task. This is the central question of organizational theory, but has not yet been well-linked to pay-setting and inequality data.
Beyond organizations, non-market interdependence can come from attempts to maintain pay norms: here the original externality workers address with a norm is a run-of-the-mill market one, but the second-order norm enforcement problem raises non-market interdependencies.
In this recent paper, I explore a third type of non-market interdependency: the circulation of tasks and scarcities across workers within a workplace.
The core idea is that when a worker holds exclusive job turf within an workplace, she is more difficult to replace than her coworkers. When employers try to simplify and deskill jobs through specialization, they unavoidably create job turf for some workers. The result is increased inequality based on workers’ different positions in an organizational task structure.
For example, if you are the only person in your department who can teach methods, you may have a bit more bargaining power than someone whose expertise is entirely redundant. Of course, the external labor market still matters here and sets bounds on bargaining power. But, possessing job turf within the workplace increases a worker’s bargaining power by making it harder to replace them with another co-worker.
I test this theory by taking a data opportunity that allows the study of, not just occupation or job title, but the actual activities that employees perform.
This granular data comes from employees of labor unions. Weird subgroup, I know. But, there seems to be no other US linked employer-employee data that includes information on year-to-year changes in the work tasks that people do. [But seriously, if you know of some, shoot me an email and let’s write a paper about them. Or, if you want to build on the ASQ paper, all data and code are posted here.]
My core research design estimates the wage effect of job turf by following the same worker, in the same job title over time, and controls for task-specific changes in local labor market demand and changes in task content. As a worker gains more job turf, her pay increases.
Moreover, the interaction between job simplification and job turf generates inequality: the workers who lose out from job simplification are not the ones who gain from the resultant job turf. These findings are robust to a series of potential objections you might be thinking about (including a fun IV looking at task changes associated with union political campaigning).
So, what’s the upshot of this? Even in an era in which traditional labor market institutions and bureaucracy is substantially weakened, non-market interdependencies among workers still matter for pay-setting.
One version of this is local, workplace-specific skill scarcities generated by the ongoing division of labor. Some workers have job turf and others don’t.
But, more generally, one way forward for sociologists of pay setting is to study the non-market, non-bureaucratic interdependencies that affect workers’ pay and inequality. The labor market and the workplace remain full of messy interdependencies. As bureaucratic constraints on pay-setting recede in importance, these remaining interdependencies will likely become increasingly important.
Nathan Wilmers. “Job Turf or Variety: Task Structure as a Source of Organizational Inequality” in Administrative Science Quarterly 2020.