Figuring out What Workers Want
In this holiday season, we hear a lot about what people want.
Most kids and many adults want presents of various sorts. Other people may want to lose weight, eat healthier, or exercise more in the new year.
All this attention to what people want reminds me that I want something too. I would like scholars who study work, occupations, and organizations, to spend more time collecting and analyzing what people want from their jobs (i.e., studying work-related preferences).
Sociologists have long recognized the importance of preferences. The rewards people get in life are related to what they want. Classic studies by Sewell and colleagues, for instance, showed that aspirations influenced status attainment. In recent years, scholars who study culture have also shown a renewed interest in preferences.
In “What People Want: Rethinking Poverty, Culture, and Educational Attainment,” Vaisey (2010) argues that poverty scholars have been too quick to dismiss the role of values in explaining the educational outcomes of the poor. Drawing on a nationally representative longitudinal data, he shows that poor and non-poor youth have different educational aspirations and that these aspirations help account for differences in educational outcomes. Vaisey proposes that scholars working at the nexus of culture and poverty need to integrate values and preferences into their theories. Scholars who study social networks have also made progress by developing methodologies to better separate preferences from the confounding influence of the opportunity structure (see Zeng and Xie, 2008).
Economists, who have historically avoided the study of preferences (based on the argument that behaviors are the best indicators of what people want), have also begun to pay attention to preferences. In “Does culture affect economic outcomes?” Guiso, Sapienza, and Zingales (2006) argue that group-level beliefs influence individual preferences, which then influence decision making and economic outcomes. In fact, economists have offered some very intriguing discussions of preferences and meta-preferences, i.e., our preference regarding which preferences we have. When sitting at the holiday feast, I may prefer to eat a piece of pie rather than fruit. I may also wish, however, that I did not have that preference: my meta-preference is to have a preference for fruit over pie instead of my actual preference for pie over fruit. Over time, tensions between preferences and meta-preferences can lead to behaviors that reflect the influence of both.
Sociologists have made notable contributions to the on-going dialogues about work-related preferences and their connections to behavior. Books like Work-lifestyle Choices in the 21st Century by Catherine Hakim, Opting Out? by Pamela Stone, and The Time Divide by Jerry Jacobs and Kathleen Gerson have all generated tremendous interest in work related preferences. Arlie Hochschild also prompted a lively debate when she suggested in The Time Bind that some parents prefer to spend many hours at work to avoid difficulties at home.
Attention to preferences in the study of work is important. People have preferences regarding all kinds of work-related issues starting with whether to seek paid employment at all. Once employed, they are likely to have preferences regarding non-standard employment arrangements, vacation time, parental leave, high performance work practices, unionization, reward systems, flexible work hours, telecommuting, nonstandard work schedules, and a host of other workplace issues. Those preferences guide their choices.
So information about preferences is vital to better understand the inequalities that result at work. They can help us better understand why some people have regular full-time jobs while others have part-time or temporary jobs. They can help us understand gender differences in the use of vacation time, parental leave, and non-standard work schedules. By knowing what workers want, we are in a better situation to uncover the structural constraints that sociology emphasizes.
In my estimation, there is a great need for conceptual and methodological development in the study of work-related preferences. To improve our understanding of work-related preferences, we must clearly define what we want to measure. When collecting information about the number of hours people prefer to work, for instance, different surveys have taken very different approaches. The wording of the questions suggests very different goals and results in very different estimates of what people prefer.
Some survey questions ask respondents what they would prefer under a very specific set of constrains. The Current Population Survey, for instance, asks:
If you had a choice, would you prefer to work:
The same number of hours and earn the same money?
Fewer hours at the same rate of pay and earn less money?
More hours at the same rate of pay and earn more money?
This question provides a clear reminder that changing works hours has consequences. Unfortunately, it overlooks the fact that changes in work hours would not have the same consequences for all workers. People who are paid a salary and those who are exempt from overtime pay may not see any increase in their earnings if they work more hours. Salaried workers may not see any reduction in their earnings if they reduce their hours. Ultimately, this means that many people who are answering the question above are reporting the hypothetical preferences they would have under circumstances that do not reflect their current situation.
Other survey questions about work hour preferences provide much different guidelines for respondents. The National Study of the Changing Workforce asks:
If you could do what you wanted to do, ideally how many hours in total would you like to work each week?
By not mentioning the potential consequences of changing one’s work hours and by using the word “ideally,” this question encourages respondents to think past the constraints they currently face and imagine a better situation. This type of information is useful for those interested in institutional change because it indicates how respondents would like to change their situations as opposed to what option they prefer given the current range of choices, none of which may be very palatable. The potential difficulty is that if many respondents report what they would prefer under unrealistic scenarios (e.g., seldom or never working but getting full pay), their answers are not very useful for understanding the constraints they currently face or for making policy.
Although both questions are about one very particular kind of preference, a careful reflection on their strengths and weaknesses brings us to central issues in the study of many work-related preferences: do we want workers to report their preferences under a set of constraints that we specify or do we want to know what they prefer given the constraints they actually face? I think we want both, and more.
Specifying constraints may be useful for developing policy, but asking respondents about their actual situations will help us understand their behaviors. Neither of these options, however, is guaranteed to reveal much about how attractive workers find the available options. Just because workers are more likely to choose a particular option in real life or in a survey does not mean they find that option satisfactory or desirable—they just find it better than the alternatives, at least under a particular set of constraints.
To better understand worker’s preferences, we need to gather information that helps put the preferences in context. We need to do a better job getting at the nuances that authors like Stone uncovered in her qualitative work, which showed that women choose to leave the labor force reluctantly because they felt they had no other choice.
It would be useful to know:
- the options workers have
- which of the options they prefer
- why they choose a particular option
- how satisfied they are with their options and choices
- what options they wish they had but do not
This type of information is difficult to gather using survey methodology, but we can certainly improve on existing approaches. Moreover, this type of contextual information is crucial for understanding the preferences that workers report and the differences between contented workers and workers who have made anguished choices or settled for the lesser evil while desperately wishing they had other options.
Furthermore, it would be useful to track people’s preferences over time. To the extent that changes in people’s preferences lead to changes in their actual situations, there is evidence that people are able to fulfill their preferences. In contrast, when job characteristics or rewards change without much connection to worker preferences or when preferences change to fit a stubborn reality, it is likely that social structures are preventing workers from getting what they want.
Gathering such data, of course, would be costly and time consuming and it would take quite a while before adequate data about preferences are available. With respect to data about work hour preferences, for instance, the US is far behind: the country has no regularly scheduled nationally representative panel surveys that include information about how many hours people prefer to work. Even if an existing panel survey like the National Longitudinal Study of Youth added questions about preferred work hours, it would take a decade or more before the available data would rival what is already present in the British Household Panel Survey, the German Socio-Economic Panel, or the Household and Labor Economics in Australia survey.
Given the difficulty of devising questions about preferences, not to mention the resources needed to collect such data, it is unlikely that the new year will bring heaps of new data about US workers’ preferences. Still, why give up hoping?
Work in Progress 