Printer Friendly

Work sharing programs: an evaluation of their use.

Work sharing programs: an evaluation of their use

Among the most innovative changes to basic unemployment insurance since its inception is short-time compensation. The programs allow workers to receive partial unemployment benefits in the event that they suffer even moderate reductions in their work hours, as long as those reductions are expected to be temporary. Employees who have their work time reduced by, say, 1 day per week might be eligible for one-fifth of their weekly unemployment insurance benefit amount. This policy differs from previous unemployment insurance regulations, under which such workers would typically be ineligible for any benefits. It is generally believed that broadening the conditions under which unemployment insurance benefits may be paid will help reduce the "prolayoff' bias of unemployment insurance and, instead, encourage employers to adopt reduced-hours strategies during periods that necessitate reducing employment levels.

This report is based on research on short-time compensation programs undertaken in response to Section 194 of the Tax Equity and Fiscal Responsibility Act of 1982. Through the Act, Congress recognized the growing number of States that have adopted short-time compensation programs as part of their overall unemployment insurance systems, and raised many important questions in response to those State efforts.1

The short-time compensation experience we will refer to occurred from mid-1982 to mid-1984 in the three States that have operated programs for the longest period of time: California, Arizona, and Oregon. The study focused primarily on the behavior of employers, although some issues that pertain more directly to employees were addressed with employee data aggregated on a per-employer basis.2

Before results of the study are discussed, three important limitations of its overall design should be stressed. The first is that the study involved only three States, each of which exhibited very low levels of short-time compensation use. The implication is that it is extremely difficult to generalize from the experiences of these States to other States that are using short-time compensation or that might use it in the future. The second is that the study did not collect data directly from employees. Thus, many issues that pertain to the attitudes and overall well-being of workers could not be addressed directly. However, evidence on some of these issues was available from information provided by employers and through unemployment insurance records. The third is that the use of a comparison-group methodology may be problematic. Both the operational status and the limited use of the work sharing programs in the study's States precluded an experimental design that would have assigned firms randomly to short-time compensation and control groups. Hence, it is possible that uncontrolled-for differences in characteristics among firms that used short-term compensation, and those that did not, may have contributed to some of the observed differences in outcomes.

Work sharing versus layoff

As noted, the primary purpose of short-time compensation is to provide firms with an alternative to layoffs during temporary downturns in their demand for labor. Many of the potential social benefits from work sharing (such as reduced labor turnover costs or increased work force productivity) derive from the ability of the program to encourage firms to substitute hours reductions for layoffs. To examine this substitution, we chose to focus on the hours spent on regular unemployment insurance and on short-time compensation by workers in our sample firms. These data were normalized by total hours employed in the fiscal 1982 base period, so as to create measures of the percentage of work time spent in these two forms of compensated unemployment. Although measuring layoffs and hours reductions with data on compensated unemployment poses some conceptual disadvantages, we believe that they are largely outweighed by the enhanced accuracy of data on compensated unemployment (because the data come from administrative records) and by the relevance of compensated unemployment to various issues of unemployment insurance financing. These data were used to examine differences in the percentage of work hours spent on compensated unemployment between firms that were short-time compensation users and those that did not participate in the program.

Although the findings exhibited fairly large State-by-State differences, three major patterns were apparent. First, participation rate in short-time compensation was very low: generally less than 1 percent of all employers in the sample States participated at any point in time, and less than 1 percent of unemployment insurance benefit claims were for short-time compensation. Second, the average firm in the short-time compensation sample appears to have continued using layoffs as the primary method of work force reduction. In no State did work sharing represent more than 25 percent of total hours on compensated unemployment among firms that used the program. Third, although evidence clearly suggests that short-time compensation did reduce layoffs (as measured by the reduced receipt of regular unemployment insurance), this substitution does not seem to have been on an hour-for-hour basis. In all of the States, firms that used short-time compensation experienced some net addition to total hours of compensated unemployment (both regular unemployment insurance and short-time compensation), although this addition was quite small in Oregon. Hence, the empirical results suggest that the actual work force-adjustment strategies adopted by firms under short-time compensation might be quite varied and complex.

The large State-by-State differences in the apparent response to short-time compensation use posed a number of questions that could not be answered satisfactorily. Whether they represented differences caused by how the program is administered, unmeasured differences among the types of firms in the various State samples, or some undiscovered methodological problem in these comparisons could not be determined conclusively given the small number of States involved and the limitations inherent in a comparison-group methodology. However, broadly similar results were obtained through a wide variety of econometric methods.

A number of other topics pertaining to layoffs and short-time compensation use are investigated in the technical report.3 Perhaps the most interesting finding from these additional investigations is that work sharing appears to have had no discernible effect on the demographic composition of layoffs undertaken by participants. That is, contrary to what has been hypothesized about the program, short-time compensation did not seem to have major "affirmative action' advantages for newly hired minority and female workers.

Unemployment insurance trust fund

Concern that widespread use of short-time compensation might have a negative impact on unemployment insurance trust funds has made many States cautious in adopting such programs and has prompted them to include in their laws special surtax provisions for firms that use the program. This concern appears to have arisen from two sources. The first is the possibility that workers who are placed on short-time compensation may have somewhat higher wages and, hence, higher weekly unemployment insurance benefit amounts than do workers who may have been laid off. Second, if short-time compensation encourages more compensated unemployment than would have occurred under a layoff-only strategy (as our results suggest), benefit payments to workers in firms that use short-time compensation may also increase.

The study results tended to support these presumptions. In all of the States, mean per-employee benefit charges (for regular unemployment insurance and short-time compensation) were significantly higher in the samples of firms that used short-time compensation. About half of the differences appear to be related to the higher unemployment insurance benefit levels for which short-time compensation recipients are eligible, and the other half can be attributed to additional amounts of compensated unemployment.

However, we also found that firms which participated in work sharing tended to have greater increases in their unemployment insurance tax rates over the study period than did otherwise similar firms who did not use the program. Although developing a precise model of how these extra tax collections should be netted against the additional benefit amounts paid was beyond the scope of the project, it was possible to provide a rough, qualitative assessment. In the short run, it is likely that short-time compensation imposed some drain on the unemployment insurance trust fund. Lags in the operations of the States' experience-rating formulas make it unlikely that the extra benefits payable under short-time compensation can be recouped in 1 or 2 tax years. Over the longer term, however, the picture is quite different. Because short-time compensation benefit charges under most of the States' current surtax provisions are more effectively experience-rated than are regular unemployment insurance benefit charges, it seems likely that any extra benefit charges would be fully recouped in the long run. Hence, concerns about the fiscal impact of work sharing might properly be focused on ensuring trust fund adequacy during temporary downturns, because the longer term solvency of the system would not seem to be imperiled by short-time compensation programs if surtax provisions are included in State laws.

Administration of the program

In adopting short-time compensation amendments to their basic unemployment insurance laws, all States attempted to restrict the concept's use to its intended purpose of substituting for layoffs during temporary downturns. In our study, we catalogued a wide variety of these provisions, and attempted to establish their effectiveness. The following general picture emerged: programs experience some startup problems but, on the whole, seem susceptible to careful monitoring and direction. We did find that some of the specific provisions of State laws do not appear to effectively ensure adherence to the basic goals of short-time compensation. (For example, constraints on minimum required work reductions do not seem to restrict minimum hours reductions to at least one averted layoff.) But these shortcomings seem relatively minor, and, in any case, the small number of States in the analysis precluded any quantitative assessment of the impacts of variations in such provisions.

In terms of unemployment insurance administrative costs, work sharing has both advantages and disadvantages relative to an equivalent work force reduction that involves only layoffs with the associated unemployment insurance collections. Initial claims-filing costs are lower under work sharing (because eligibility determination is simpler), and cost associated with unemployment insurance "work test' monitoring do not arise. However, relative to an equivalent level of layoffs, short-time compensation entails processing a much greater number of weekly benefit claims. Our analysis (which was constrained by a relatively meager amount of administrative cost data) suggested that, under current circumstances, the costs associated with the greater volume of claims tended to dominate the lower per-claim costs. However, in interpreting this conclusion, it is important to keep in mind the relative newness of short-time compensation programs. It is quite possible that the administrative cost disadvantages of them may decline over time as experience with the programs accumulates.


1 The questions focused on the effects of short-time compensation on contemporaneous and subsequent layoffs, unemployment insurance tax rates, the integrity of the unemployment insurance trust funds, and the net benefits to the various affected parties.

2 The analysis was based on survey and unemployment insurance records data covering 1,050 firms spread across the three States. Approximately half of the sample had used short-time compensation during the study period, and the other half did not. Nonusing firms were carefully matched to the program sample on many observable dimensions, and the actual analysis was based on regression techniques that controlled statistically for possible remaining sample differences.

3 Stuart Kerachsky and others, "An Evaluation of Short-Time Compensation Programs: Technical Report" (Mathematica Policy Research, Inc., Princeton, NJ, December 1985), prepared for the Employment and Training Administration, U.S. Department of Labor.
COPYRIGHT 1986 U.S. Bureau of Labor Statistics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1986 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kerachsku, Stuart; Nicholson, Walter; Cavin, Edward; Hershey, Alan
Publication:Monthly Labor Review
Date:May 1, 1986
Previous Article:Short-time compensation: assessing the issues.
Next Article:Short-time compensation: the AFL-CIO perspective.

Related Articles
Evaluation: what to measure and why?
Truck Safety: Share the Road Safely Program Needs Better Evaluation of Its Initiatives.
Is your research improving health services?
Funders want more evaluations but not ready to pay the cost.
Anthony N. Maluccio, Cinzia Canali and Tiziano Vecchiato (Eds.) Assessing Outcomes in Child and Family Services: Comparative Design and Policy Issues.
University-community partnerships; universities in civic engagement.
Other principals' departures announced.
Communicating evaluation findings from offender programs.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters