Why does part-time employment increase in recessions?

Structural part-time employment

By definition, part-time work differs from full-time work in terms of the number of usual hours worked. A common definition is to consider as part-time workers those individuals who work fewer than 35 usual hours per week. This is the definition used by the US Bureau of Labor Statistics and the one used throughout this article. Based on this definition, part-timers work, on average, about half the number of hours of full-timers (Figure 1). Some statistical agencies, for example, Eurostat, also use individuals’ self-reported part-time employment status, or a 30 hours cutoff, to define part-time workers. In general, using different definitions of part-time work affects the overall levels of part-time employment, but provides a similar qualitative picture of the various roles played by part-time employment in the labor market.

In addition to differences in hours worked, the composition of part-time employment is quite different from that of full-time work. This follows from the specific advantages (relative to full-time employment) that part-time work offers to certain workers and employers. For women, the young, and the old, a part-time schedule offers the opportunity to combine work in the labor market with house and caring activities, education, or leisure. Consequently, in most countries these demographics account for the largest fraction of part-time employment. This type of part-time employment is predominantly voluntary, since individual preferences (related to the life cycle) lead workers to choose part-time over full-time work. In the absence of part-time arrangements, these workers would face the choice between working full-time and not working at all. Therefore, many countries actively promote part-time work as way to increase labor force participation.

From the employers’ perspective, the main advantage of part-time work is the increased flexibility it affords them in managing irregular demand patterns. In industries where opening hours are either very long or very short, and during which demand is subject to predictable peaks, it is often more cost-effective for firms to employ a part-time workforce. Accordingly, part-time work is used much more intensively in the retail trade, leisure and hospitality, and education and health services, and much less so in construction and manufacturing. Even though it appeals more to certain workers and employers, part-time work is a pervasive form of employment across the labor market. In many industries, occupations, and other forms of employment (e.g. temporary and self-employment), the share of part-time workers is considerable. Similarly, part-time employment is present in most segments of the employed population defined by differences in age, education, and gender.

In choosing part-time over full-time work, both workers and firms should weigh the relative benefits of part-time employment against its relative costs. In addition to the structural factors mentioned in the previous paragraphs, labor market regulations are likely to affect the relative costs of part-time work in important ways. In most developed countries, part-time workers are entitled to the same working conditions and pay (on a pro-rata basis) as similar full-time workers. However, despite the recent progress made in several countries to enact equal treatment policies for part-time and full-time workers, in practice, these legal provisions are difficult to enforce. Therefore, working part-time is likely to entail worse working conditions and lower access to social security and other benefits compared to full-time work.

A raw comparison of the average hourly wages of part-time and full-time workers delivers a substantial part-time wage penalty in most countries. However, empirical studies find that the part-time wage penalty is largely explained by differences in worker skills (education and experience) and job characteristics (industry and occupation) [1], [2]. Consistent with these findings, in normal times (i.e. non-recessionary), part-time work is overwhelmingly voluntary.

Cyclical part-time employment

A less well-known feature of part-time employment is the fact that its share of total employment varies with the business cycle. In recessions, part-time employment usually goes up and during expansions it goes down again. Recessions are periods of long-lasting reductions in aggregate economic activity, which prompt workers and firms to adjust to the new economic conditions. These adjustments can take many forms and usually differ across types of workers and firms. Some firms face lower demand and are forced to reduce their costs by firing their employees, or lowering their wages and/or hours worked. Some workers who are outside the labor force in normal times may be forced to join it in a recession to compensate for the reduction in household income due to a spouse’s dismissal. In a typical recession, the interaction of these various decisions results in a decrease in the number of employed workers, as well as a reduction in the average hours worked among employed workers.

During the Great Recession, employment not only dropped in most countries, but also the share of workers employed full-time. In the same way that workers who lose their job usually remain available and willing to take a job, most workers who suffer a reduction in working hours are available and willing to work more hours. Consequently, cyclical reductions in employment largely translate into increases in unemployment, and increases in part-time employment are chiefly driven by involuntary part-time work. Unlike unemployment, the definition of involuntary part-time employment is not uniform across countries. In most labor force surveys, involuntary part-time employment refers to individuals who work part-time because they cannot find a full-time job. In the US labor force survey (the Current Population Survey), respondents can also indicate “slack business conditions at their current employer” as the main reason for working part-time. The extra detail available in the US data shows that, in a typical recession in that country, the number of part-time workers stating slack business conditions plays a predominant role in the increase in overall involuntary part-time work [3], [4], [5].

Why does part-time employment increase in recessions?

Given that the incidence of part-time employment is much higher among specific groups of workers and firms during recessions, one hypothesis is that cyclical changes in the composition of employment are behind the recessionary increase in part-time employment. For example, if construction and manufacturing employment drops very quickly at the onset of recessions, whereas employment in retail trade and services is more resilient, this will result in an increase in the share of workers employed part-time. The data show, however, that this hypothesis only accounts for a small part of the story [5].

Figure 2 illustrates this point using data from a selection of EU countries, the UK, and the US. For each country, it reports the actual peak-to-trough change in involuntary part-time employment during the Great Recession, alongside counterfactual peak-to-trough changes that keep the composition of employment fixed during the recession. If the increase in the involuntary part-time employment share is explained by changes in the composition of employment, the counterfactual shares should be significantly smaller than the actual one. Figure 2 dispels this possibility. Consider the case of Austria, described in the first row of Figure 2. The first column (“Actual change”) indicates that the involuntary part-time employment share increased 0.48 percentage points (pp) in that country. The next column (“Age”) indicates that this change would have been very similar (at 0.45 pp) had the distribution of age among employed individuals remained fixed to what it was at the recession’s peak. Comparing the value in the first column with the remaining columns, which fix the distribution of employment by sex, education, industry, and all dimensions simultaneously, leads to similar conclusions. Applying the same logic for the other rows delivers the same conclusion. In other words, this evidence suggests that the recessionary increase in involuntary part-time employment is not driven by the cyclical reallocation of labor from less to more part-time intensive segments of the labor market.

An alternative hypothesis is that part-time employment increases due to downward adjustments in hours worked among employed individuals. In other words, the drop in economic activity forces firms to reduce labor costs, and many choose to do so by reducing the average number of hours worked per employee. If the costs of hiring and firing employees are high, employers and employees may find it optimal to respond to a temporary fall in demand by reducing working hours. Recent research finds strong support for this hypothesis in the UK and the US labor markets [3], [5].

To assess this hypothesis, recent research studies the dynamics of part-time employment through changes in the rates at which workers move across labor market states (e.g. from unemployment to full-time work, or from part-time work to non-participation). The data show that unemployment and non-participation play only a minor role in the aggregate reallocation of workers from full-time to part-time hours. This reinforces the idea that workers are not moving from “good” full-time jobs in contracting sectors/firms to “bad” part-time jobs in expanding ones with an intervening spell of non-employment. Instead, in the US and the UK, the onset of recession leads to a jump upwards in the probability of full-time workers being employed part-time in the next period (i.e. the next month or quarter), and a drop in the probability that part-time workers become full-time employed in the next period. Together, these two changes account for the lion’s share of the increase in part-time employment in a typical large recession in the US and in the Great Recession in the UK. Figure 3 illustrates this pattern in the past two decades in these countries, focusing on transitions between full-time and involuntary part-time employment. The two lines exhibit a very clear counter-cyclical pattern, increasing sharply as the recession sets in and falling again during the recovery. The much higher levels of the US transition probability are due to the higher levels of US labor turnover, as well as differences in how involuntary part-time work is recorded in the Current Population Survey.

A key fact uncovered by recent research is that a very large fraction of worker transitions between part-time and full-time work occur at the same employer and usually entail sizable changes in working hours (on average around one and a half working days per week in the UK and the US). This is the case both in normal times and during recessions [5]. In other words, for a large fraction of employment relationships, working hours can adjust very quickly and by a sizable magnitude. This picture contrasts with the structural view of part-time employment, according to which worker preferences and employer technology dictate the part-time/full-time status of the job. Instead, it brings to light the notion that part-time employment is an operative margin for workers and firms to temporarily adjust the terms of the employment relationship in response to shocks. Consistent with this view, the elevated levels of high-frequency reallocation between part-time and full-time employment affect not only those segments of the population that are more prone to take on a part-time jobs (like women, the young, and the old), but also prime-aged men (those aged 25 to 54).

How bad is cyclical involuntary part-time work?

Since part-time work usually entails worse employment conditions and lower pay, it seems natural that policymakers should worry about the very large increase in involuntary part-time work observed in recessions. However, recent research based on the US labor market suggests that involuntary part-time employment is a relatively benign labor market outcome when compared to unemployment. The comparison to unemployment (instead of full-time employment) is the relevant one to make. Recessions always entail some form of labor adjustment—indeed, some labor adjustment is beneficial for the short- and long-term performance of the economy—so a comparison with full-time work is of lesser practical importance. In other words, the relevant trade-off faced by employers and employees is the one between layoffs and reductions in hours, not between reductions in hours and no adjustment.

In comparing the effects of worker relocation from full-time work to unemployment versus the relocation to involuntary part-time work, consideration should be given to the two main differences between these alternative labor market states. The first one to consider is the reduction in income relative to full-time employment. In the US, reallocation from full-time to part-time work or unemployment entails similar reductions in income, of about 12% to 14% of annual income [6]. The reason for this is found in the design of existing public income insurance programs. Most US states have unemployment insurance systems in place that provide income to workers who lose their job, as well as to part-time workers who are searching for a full-time job and/or are working fewer hours at their current employer due to slack demand conditions. On top of their unemployment benefits, involuntary part-time workers earn some labor income, but the amount of benefits paid is adjusted to the level of income. Specifically, any labor income earned over a specific level is subtracted in full from the maximal amount of full-time unemployment benefits the worker is entitled to, based on their past contributions. In addition to partial unemployment insurance, uninsured workers who become employed part-time involuntarily may receive income compensation as part of short-time compensation schemes, or through the Earned Income Tax Credit (EITC). However, these programs cover a much smaller fraction of the employed population. Short-time compensation schemes exist in fewer than half of US states and the take-up rate by involuntary part-time workers in those states is very small [7]. The EITC is designed to offer income support to low-income working families with children and, in practice, only a very small fraction of involuntary part-timers is eligible for the EITC [7].

A second relevant aspect concerns the duration of spells of unemployment and involuntary part-time employment. In this respect, involuntary part-time work differs quite dramatically from unemployment. On average, involuntary part-time spells are shorter than unemployment spells and, more importantly, they are much more likely to lead to a return to full-time work. Amid the Great Recession in the US, the probability of an involuntary part-time worker returning to full-time work was close to twice as high as that faced by an unemployed worker [7]. For forward-looking individuals, a quicker return to full-time employment implies a higher expected future income and, therefore, is likely to lead to a lower reduction in current consumption. Consistent with this view, recent US estimates of changes in food consumption following reallocation from full-time employment to involuntary part-time employment and unemployment show that the drop is much higher for unemployed workers. For a full-time worker who moves to unemployment, their food consumption is likely to change, on average, between −8% and −10%, whereas the change in food consumption for a full-time worker who moves to involuntary part-time employment is basically zero [6]. Since food is an essential consumption item, these differences are probably smaller when looking at the effects on overall consumption. In any case, this should not alter the fact that the overall consumption effects of involuntary part-time work are lower than for unemployment.

What explains the elevated persistence of involuntary part-time employment?

Although UK and US involuntary part-time employment shares have been steadily declining, many years into the recovery they remain clearly above their pre-crisis levels. This contrasts with a full recovery in unemployment observed in both countries, as well as with the experience of past US recessions, when unemployment and involuntary part-time employment recovered at a similar pace. Taken at face value this suggests the prevailing increase in involuntary part-time employment reflects structural, as opposed to cyclical, factors. There is some US evidence suggesting that persistence in involuntary part-time work during the recovery is due to changes in the industry composition of employment [4], as well as evidence that the composition of involuntary part-time employment is shifting toward workers who cannot find a full-time job and away from workers on slack business conditions [3], [8]. On the other hand, the lack of signs of a strong recovery in both countries suggests elevated part-time employment may be a sign of weak economic conditions.

An alternative explanation in the US context is the adoption of the Patient Protection and Affordable Care Act (ACA) in 2010, which mandated employer-sponsored health insurance coverage for firms with 50 or more full-time equivalent employees. Since the mandate only pertains to employers’ full-time employees, firms may adjust to the legislation by increasing the share of part-time jobs. Indeed, there is evidence that similar state-level legislation adopted in Hawaii and Massachusetts led to an increase in part-time employment among low-skilled workers [9], [10]. Due to delayed application of fines on employers until 2016, decisive evidence on the effects of the ACA is still lacking. While this explanation may have some traction in the US, it cannot explain developments in the UK.