Abstract

Based on two sets of data—the National Bureau of Economic Relations (NBER) Shared Capitalism Project surveys of employee stock ownership plan (ESOP) companies and the national General Social Survey—we find that employees in ESOP companies have significant holdings of employer stock. While salaried workers in ESOP companies have larger account balances than hourly workers do, hourly workers in ESOPs do significantly better than hourly workers in general, both in terms of ownership and earnings. More broadly, the median pension wealth of ESOP participants is over four times higher than the median household pension wealth. ESOP retirement plans do, however, remain too concentrated in employers' stock.

Cheerleaders for the "ownership society" tout the growing share of U.S. households owning stock—from 31.7 percent in 1989 to 51.9 percent in 2001 counting both direct and indirect ownership (Wolff 2004, Table 12b). What is less often advertised is that as of 2001 the bottom 90 percent of households own only 23 percent of all stock and just 12 percent of all directly held stock, which confers direct control (voting) rights on stockholders (Wolff 2004, Table 13a; Kennickell 2003, Table 10). Only 30 percent of households in the bottom 90 percent of the wealth distribution owns (directly or indirectly) more than $10,000 of stock (Wolff 2004, Table 13a). If the ownership society is defined in terms of households' ownership stake in the U.S. economy, the huge majority of American households are decidedly minority owners.

These statistics on the concentration of stock ownership are reflected in the data on pension wealth as well. Thirty-four percent of households have no pension plan (Wolff 2005, Table 5), and "more than one-fifth of all households nearing retirement (those between the ages of 56 and 64) had no retirement savings other than Social Security" (Weller and Wolff 2005, 2). In 2001 average pension wealth (in defined benefit and defined contribution plans) was $94,800, but the median—the pension wealth of the typical household—was only $10,900 (Wolff 2005, Table 11). This paper assesses the potential of employee stock ownership plans (ESOPs) for increasing both workers' pension wealth and their ownership stake in the companies they work for.

Employee Stock Ownership: Forms and Financials

Blasi, Kruse, and Bernstein (2003, Appendix C) calculate that in 2002 there were 24.1 million participants in 11,561 pension plans that held company stock.¹ About 8.2 million (34 percent) of these participants were in ESOPs, and these held 59 percent of all company stock in employee pension plans. In all, it is probably safe to say that at least half of all company stock owned by employees is held in ESOP retirement accounts (Rosen 2005). ESOPs were created as a form of defined contribution pension plan by legislation attached to the Employee Retirement Income and Security Act (ERISA) of 1974. Favorable tax treatment of income from the sale of company stock to ESOPs and of company profits contributed to ESOPs has encouraged their rapid growth and made them the most common form of employee ownership in the United States.

Several studies of the wealth effects of ESOPs have been conducted in recent years. These include (1) a census of Washington State ESOPs (Kardas, Scharf, and Keogh 1998), which found median pension assets per participant of $31,600 (versus $5,400 for a matching sample of non-ESOP control companies); (2) a 2005 study of Ohio companies, which found median ESOP account balances of $30,000 (cited in Rosen 2005); (3) a 2005 census of Massachusetts ESOPs, which reported average assets per participant of $56,200 (Mackin 2005); and (4) a survey of sixteen S corporation ESOPs, which found median employee account balances of $75,000ø$100,000 (Rosen 2005).

Here we develop new estimates of employee stock ownership and pension assets based on two large-scale employee surveys: the 2002 national General Social Survey (GSS), which provides a representative sample of 1,120 private sector employees, and the National Board of Economic Research (NBER) Shared Capitalism Research Project, which includes a sample of 6,000 employees in nine ESOP companies.²

A Profile of Nine ESOPs

There are reasons to believe that employee stock ownership might have positive effects on workplaces and company performance, and the evidence, though mixed, broadly supports this view on some important measures (Kruse 2002; Kruse, Freeman, and Blasi 2006). Table 1 provides a first look at workers' ownership stakes in the nine ESOP companies in the NBER study. These firms vary in size from a few hundred to thousands of employees. They are spread across service and manufacturing industries, and they vary in the length of time that their ESOPs have existed (from four to twenty-four years) and in the percentage of the company's stock held in the ESOP.

ESOPs are broad-based defined contribution pension plans, which normally include all full-time (and often also permanent part-time) employees meeting basic requirements of age, hours of service, and months of service, with limited exceptions allowed by ERISA) Self-reported participation rates among the nine companies in the NBER study vary from 64 to 95 percent, with an average of 84 percent. It may be worth noting that the three companies with the highest participation rates (all around 95 percent) have by far the highest average account balances—ranging from $129,000 to $257,000. At the other end of the range, the lowest average account balance is just $5,400. In all, the average account balance of all ESOP participants was $104,600, but including employees without ESOP accounts and participants with zero balances reduces the mean account balance of all employees to $81,500.

The ratio of mean to median account balances is an indication of the dollar "distance" between the stock holdings in the largest accounts versus the typical account. This varies from 1.1:1 to 3.6:1 and is 3.5:1 for all ESOP participants in the data set. This higher ratio for all employees reflects the effect of interfirm differences in average account balances as well as the intrafirm inequality reflected in the company-specific ratios.

All but one of the NBER companies had 401(k) pension plans as well as ESOPs, and almost two thirds of the employees of these ESOP companies had 401(k) retirement plans—one of which included company stock in the

401(k). Table 2 shows the value of total pension assets held in all retirement plans and the value of employer stock held in these plans. Mean pension assets for all employees of these ESOP companies is $105,900, and the mean for those who own some employer stock is $130,300. The mean value of employer stock held by employee owners is $90,600, and an average of two thirds of employees' pension assets are in employer stock (though this percentage varies widely among companies from 25 percent to 100 percent). These figures suggest that while most workers in ESOPs have other diversified retirement plans, the bulk of their retirement assets remain invested in their employers' stock. Diversification of pension assets would seem to remain an important issue for many ESOP participants.³

The Distribution of Stock Ownership and Earnings within Firms

Table 3 compares the employees of the NBER ESOP companies with the GSS national survey of employees (or the subset of those who work for private for-profit companies). The GSS data indicate that about 25 percent of private sector employees own company stock, including 34 percent of salaried employees and 19 percent of hourly workers. Of course, the percentage of owners is much higher for employees of ESOP companies, and, more importantly perhaps, the ownership gap between salaried and hourly employees is much smaller (in relative terms).

Average holdings of employer's stock by salaried employees in ESOP companies is $130,500, and the average holdings of salaried ESOP participants is $149,800. For the wider economy represented by the GSS sample, salaried employee-owners have an average of $170,200 in company stock. But hourly employee-owners in the GSS hold an average of only $24,000 of their employers' stock versus $47,100 for hourly employee-owners in the NBER companies.

Employer stock in employee ESOP accounts is purchased by the ESOP and given to employees. (Employer stock acquired in other plans may be paid for by the employer or by the employee.) However, the formal accounting rules might not reflect the true incidence of the cost of employer stock contributions if those contributions substitute for other forms of compensation. Thus, employees with ESOPs might pay for their company stock with lower wages than they would otherwise earn. Comparison of the earnings of employees in the NBER ESOPs versus the GSS national sample (bottom of Table 3) shows that ESOP company employees' earnings are 16 percent higher than earnings of the (private, for-profit sector) employees in the GSS ($42,745 versus $36,950). The earnings advantage to ESOP company employees is even higher for hourly employees, whose earnings are 33 percent higher than those of hourly employees in the national sample ($32,956 versus $24,829).

Employer stock in ESOPs is allocated according to legal requirements that make it no less equally distributed than taxable (W-2) pay (with a cap of $210,000 on eligible earnings that limits allocations to top management). But what about the distribution of pay in employee-owned companies? Hourly workers' earnings average 53 percent of salaried workers' earnings in the ESOP companies compared to 44 percent for the wider economy (bottom of Table 3). Hourly workers in ESOP companies get lower pay and less stock than salaried workers, but they do significantly better than hourly workers in general, with respect to both earnings and ownership.

Table 4 takes a closer look at the distribution of annual earnings in ESOP companies and the general economy. Here again we see that mean and median earnings are higher in the ESOP companies (column 1) than in the wider economy (column 2). Moreover, the earnings ratio between the highest (90th percentile) and lowest (10th percentile) paid employees is more than two and a half times higher in general than it is in ESOP companies (8.95 versus 3.48). In terms of earnings, for whatever reason, less-skilled (hourly) workers appear to do much better in ESOP companies than elsewhere.

Pension Wealth

Table 5 reports the pension wealth of participants in ESOPs, 401(k)s, and other plans in the NBER ESOP companies compared with employer stock ownership in the GSS sample and with Edward Wolff's (2005) estimates of household pension wealth in defined benefit and defined contribution retirement plans. GSS employee-owners (many of whom are undoubtedly in ESOPs) have 10 percent more employer stock, on average, than employee-owners in the NBER ESOP companies, but their median holdings of employer stock are only 29 percent as much as ESOP participants' median holdings. A comparison of all plan assets of

workers in the NBER data with Wolff's estimates of household pension wealth shows a 37 percent higher average pension wealth for ESOP participants and an over four times higher median pension wealth.⁴

Finally, we recall Wolff's calculation that the bottom 90 percent of households owns just 23.1 percent of all stock. A similar calculation for employees of the ESOP companies in the NBER data (both participants and nonparticipants) finds that the bottom 90 percent owns 34.5 percent of the company stock and 39.8 percent of all assets their retirement plans.

Conclusions

This study of employer stock ownership in ESOP companies and in the wider economy indicates that average ESOP balances are substantial, compared to average levels of company stock ownership in general. While salaried employees do much better than hourly employees in these ESOPs, hourly employees do significantly better in ESOP companies than their counterparts do in the wider economy. Moreover, there is no evidence that employees with ESOPs (either hourly or salaried) pay for their company stock in lower wages. Quite the contrary—earnings are significantly higher in ESOP companies than in general, especially for hourly employees.

Finally we note that, while far from adequate from the standpoint of financing retirement, the median pension wealth of ESOP participants is over four times higher than the median household pension wealth (Table 5) and that company stock ownership in ESOPs, while highly concentrated, is considerably less concentrated than stock ownership in general. If all employees worked for ESOP companies like those in the NBER study, the distribution of stock ownership would be considerably more equal than it is.

Acknowledgments

This research is supported by a grant from the Russell Sage Foundation and the Rockefeller Foundation. We thank Douglas Kruse for his invaluable service in compiling the employee survey data used in this study.