DECLINING UNIONIZATION: DO FRINGE BENEFITS MATTER?

INTRODUCTION AND REVIEW

Private sector unionization in the U has lengthy been declining. After reaching a high mark of 356 percent of the civilian workforce in 1954 union density has fallen steadily, to 98 percent in 1997 Seminal studies by dint of Ashenfelter and Pencavel [1969], Pencavel [1971] and Bain and Elsheikh [1976] have spawned a vast literature in which researchers try to find to identify factors responsible for the decline. Factors in the decline include changes in the constitution of the economy, management opposition to organized labor, and other socio-political factors.1

The make of compensation also has changed dramatically above time. In 1948 non-wage benefits accounted for nearly 45 percent of total compensation; by the agency of 1994 that ratio had more than quadrupled, to 19 percent [U Bureau of Economic Analysis, 1998] Despite this correlation of incidents illustrated in Figure 1, there has been no analysis whether the composition of pay has affected organized labor.

Benefits take many forms, for example, health care insurance, pension funding, vacation pay, or employer payments for social security and unemployment insurance. The costliest benefits to provide are paid leave, health insurance, and social security contributions [U Bureau of Labor Statistics, 1997; Lettau and Buchmueller 1999] The first sum of two units are provided voluntarily; social security is mandated by dint of law. Most fringe benefits are paid voluntarily, with mandated benefits constituting roughly alone one-third of fringe payments.2 This application of mind examines whether there is a relationship between benefits and unionization.

Union members receive greater non-wage benefits, upon average, than nonunion workers (see Freeman and Medoff [1984]) Moreover, organized labor has bargained and lobbied for welfare improvements, like shorter working hours, compensation for workplace injuries, and unemployment insurance. Many improvements sought by means of organized labor, in fact, have been institutionalized by dint of statute. Ironically, this legislative succes might undermine support for organized labor.

Some scholars have argued that governmental mandates can replace a certain quantity of union functions. This "substitution hypothesis" was advanced by dint of Neumann and Rissman [1984], who reported that union membership is inversely related to management workplace protections and social welfare programs.3 According to Bennett and Taylor [2001 p 261] government-mandated benefits restore the "scope of issues that may be addressed at the bargaining table." with equal reason if legislation makes certain benefits universal, unionization could suffer4 In fact, Fiorito [2001] lay the foundation of an inverse relationship between voluntary provision of benefits and workers' intentions to devoted for union certification.

Some benefits, like health insurance, can provide tax breaks for the one and the other firms and workers, so the one and the other sides have reasons to favor them. on the contrary employers may have another incentive to tilt the compensation mix towards fringe benefits. Wages and fringe benefits are suspended during a strike. Wages can be partially replaced, by dint of union strike funds or if workers take temporary piece of works elsewhere. But temporary jobs typically provide no benefits, meaning health care coverage is not to be found during a strike. As health care has become increasingly expensive, workers may be reluctant to jeopardize their coverage through risking prolonged industrial action.5

As fringe benefits have become more universal, that may have remov issues from potential collective bargaining. And as fringe benefits have become a more prominent feature of compensation, support for organized labor may have weakened.

This article reach outs research on the decline of organized labor by the agency of analyzing the impact of non-wage benefits upon unionization. Specifically, we expand Pencavel's [1971] econometric specification to estimate in what way the growing role of benefits has affected union density. Thus, our inquiry documents whether the "substitution hypothesis" can be amended. Using time-series observations for 1948-1997 we examine whether unionization nationwide has been influenced through the changing composition of pay. With cros sectional data for 1983-1996 we also analyze the repercussions for organized labor across states.

BENEFITS AND UNIONIZATION: NATIONWIDE ANALYSIS

Data and Sample

Prior to World War II non-wage benefits were negligible. During the war pay be frozens were imposed, so benefits became a means of increasing total compensation. A consistent fringe benefits series for the economy reach outs back to 1948; the latest information we have is for 1997 Thus, our sample overspreads the fifty-year period 19481997. In our regression analysis, we use FRINGE as an explanatory variable, which shows the percentage of total compensation accounted for by means of non-wage benefits (see appendix for all data sources).

The hanging variable is UNIZ, the fraction of the private, non-farm labor force that is unionized. Previous empirical studies have used different data to exhibit this series. A popular series has been erected by Troy and Sheflin [1985] which is based upon the financial reports filed through organized labor under the Landrum-Griffin Act. For each union, they divided by capita revenue by the organization's by means of capita dues rate to sum up average annual, full-time, dues-paying members. Summing above all unions yields the overall horizontal of membership. Unfortunately, the Troy and Sheflin series was discontinued after 1982