New NCEO Estimates on the Growth of Employee Ownership

For anyone interested in employee ownership, the first question is always how many plans there are. Unfortunately, the best we can do is to provide a very rough estimate of the number of plans and participants. As of the end of 1997, we estimate that that are 11,100 ESOPs and stock bonus plans, about 2,000 401(k) and profit sharing plans primarily invested in employer stock, and at least 3,000 stock option plans that provide options to 50% or more of their work force. We estimate that there are about 7.7 million participants in ESOPs and stock bonus plans with assets worth about $400 billion, about 2 million participants in 401(k) plans with assets worth about $250 billion, and about 6 million participants in broad-based stock option plans (it is not possible to estimate a value for options, however).

These numbers should be viewed with considerable caution. They represent only an educated guess. For reasons to be explored in detail in this article, it is simply not possible to generate data that are current and reliable.

What the Numbers Tell Us About the Growth of Employee Ownership

The most remarkable development in employee ownership has been the extraordinarily rapid growth in broad-based stock option plans. Almost unknown a decade ago, they are by far the fastest growing segment of employee ownership and show no sign of slowing down. Based on various surveys, it appears at least 10% of all public companies (and an unknown number of closely held ones) provide stock options to at least half their employees. Based on a conservative estimate of the number of participants in these plans, we believe that six times as many employees now are receiving options as were in 1992.

We expect these numbers to continue to grow for a number of reasons. First, most knowledge-intensive businesses are now offering options to most employees, and this sector of the economy is expanding. Second, these companies are the fastest-growing companies in the economy, so the number of participants will increase even if the number of plans does not. Finally, the concept is now spreading to other than knowledge-based businesses.

On the other hand, ESOPs are a mature phenomenon. As the total number of plans has grown, the number of new plans needed each year to offset terminations becomes larger. The ESOP termination rate has held at a steady three to four percent of total plans over the last 15 years. With 11,000 plans, that would mean there would have to be 440 new plans just to stay even. In the early part of this decade, only 400 to 500 new plans were set up each year. In recent years, that number has almost doubled, providing new growth. However, most of the growth has occurred in smaller plans, while some very large plans set up in public companies in the 1980s are now being terminated. These plans were set up to capture tax benefits no longer available or to prevent hostile takeovers. Tax and accounting rule changes have made these plans less attractive, so at least some plans are terminated when the loans are paid.

Does all this mean ESOPs are a stagnant phenomenon? Hardly. First, most of the public plans were of little significance to employees. The ESOPs usually just constituted a match to employee 401(k) plan deferrals. In most cases, the company was matching in stock before the ESOP and will continue to do so after the ESOP is terminated. All that has changed is the delivery mechanism. Second, many plans in closely held companies are terminated because the company is sold, usually at a considerable profit. This shows up as a decline in the number of ESOPs to be replaced by a new plan, a statistical zero effect. The net effect of this on employee wealth, however, is far from zero. Finally, equilibrium is a state all growth phenomenon must eventually reach. The New York Times has probably been at a circulation equilibrium for many years, but to say that it is thus a stagnant business phenomenon would be very misleading.

How We Arrived at the Estimates

Otto von Bismarck advised against seeing how laws and sausages were made. The weak of stomach may want to skip this section as well.

Department of Labor Data and Its Problems: For ESOPs, we would like to rely on Department of Labor summaries of Form 5500 reports. (Form 5500 is a tax form that companies sponsoring an ESOP or other such plan file with the Internal Revenue Service every year.) Unfortunately, the latest 5500 data are always four years old. Even worse, the estimate for the number of plans is based on annual reports from companies with over 100 participants and (here is the rub) a 10% sample on one year's filings from smaller plans (these plans only have to file every three years). Projections for 1994 were made based on just 252 plans for about 25 times that many actual companies. The result is that year-to-year results vary unpredictably. Second, the data are based on plan self-reporting and appear to riddled with errors and inconsistencies. Things are called ESOPs that we know cannot be; ESOP boxes are not checked when they should be. The 401(k) and profit sharing plan data are more reliable, but dated. We use the dated data because we do not know how to estimate recent trends.

The 5500 data inconsistencies for ESOPs are especially troubling. The data report that of 9,226 plans called ESOPs in 1993, 7,874 were "profit sharing and thrift-savings" plans; in 1994, the DOL said that of 8,736 ESOPs, 5,996 were in these categories. Only 998 plans were "stock bonus" plans in 1993, while the number jumped to 1,758 in 1994. Other categories, such as IRAs or annuities (73 plans in 1993 and 54 in 1994) seem even odder. Similarly, the DOL's report of the number of leveraged ESOPs indicates there were 1,060 in 1993 and 1,753 in 1994, even though the total number of ESOPs reportedly declined.

The problem with these data is twofold. First, the jumps from one year to the next make no sense. Nothing happened from 1993 to 1994 to explain why there would such a dramatic shift in categories. Second, the categories themselves make no legal sense. An ESOP is defined as a stock bonus plan that can borrow money. An ESOP cannot be a profit sharing plan, thrift plan, IRA, or other annuity plan.

Doug Kruse, a professor at Rutgers who is arguably the country's leading expert on profit sharing and related plans, as well as 5500 data, suggests an explanation. It could be, Kruse speculates, that companies that have, for instance, a 401(k) plan also have an ESOP. The DOL procedures find the 401(k) plan first, then put the ESOP in with these plans. So the data would be saying that there are 5,996 companies with profit sharing and thrift plans that also have ESOPs. This would be consistent with observation of most ESOP experts that most ESOP companies also have a 401(k) plan. Efforts by the NCEO to get a DOL explanation failed.

The problem with the leveraged ESOP numbers, however, seems to have no explanation. The estimates are obviously wrong. According to everyone who works with ESOPs, most plans are leveraged. The undercounting on this issues seems just a mystery.

IRS Data: The next source is IRS determination letters. Plans should, but do not have to, file for such letters when they are started or terminated. The most recent data are from letters issued in 1996 (most plans would have filed for these in 1995). For the early 1990s, however, these letters were not issued to ESOPs because of regulatory delays; in recent years, they have become more reliable, but still have a number of problems. We use these numbers for 1995, then guess at the trends for the ensuing years based on what we hear from plan providers. Plan asset value estimates are based on DOL numbers (holding our breath), then updated based on changes in equity values in general over the ensuing years.

For stock options, we rely first on an actual enumeration of plans we have identified from a major research project now underway. Names come from providers and a clipping service. We then add up the number of employees in the companies. Because most of these are public companies, and are much larger than the closely-held companies offering broad options, we can assume that the lion's share of employment will be with these companies. That gives us some confidence in the number of employees covered (albeit providing a conservative bias), but then we have to guess at the number of closely held small companies providing broad options.