Treasury Releases OTA Paper on Corporate Choice of Organizational Form

Treasury has released a copy of a paper from its Office of Tax Analysis on taxes and corporate choice of organizational form. In the paper, authors Robert Carroll and David Joulfaian examine the effect of the Tax Reform Act of 1986 on corporations' decisions to organize as C corporations or S corporations. After analyzing the corporate income tax returns of many corporations before and after TRA '86, the authors conclude that TRA '86 exerted a "statistically significant influence" on a firm's choice of organization form. In addition, the authors say, firms that switched to S corporation status after the passage of TRA '86 grew about 20 percent faster than the general corporation population.

====== FULL TEXT ======

TAXES AND CORPORATE CHOICE

by

Robert Carroll and David Joulfaian

OTA Paper 73 October 1997

OTA Papers and Briefs are circulated so that the preliminary findings

Office of Tax Analysis

ABSTRACT

[1] This paper investigates the determinants of the corporate choice of organizational form and the effect of this choice on firm performance. We analyze the corporate income tax returns of a large number of corporations before and after the Tax Reform Act of 1986 and determine how the substantial changes in tax rates associated with this tax reform affected corporations decisions to organize as C corporations or S corporations, which are taxed only at the personal level. We find that tax rates exert a statistically significant influence on a firm's choice of organizational form with an implied tax rate elasticity of 0.20 percent. Our results also show that firms that switched to S corporation status after the Tax Reform Act of 1986 grew about 20 percent faster than the general corporate population.

TAXES AND CORPORATE CHOICE OF ORGANIZATIONAL FORM

I. INTRODUCTION

[2] Following the Tax Reform Act of 1986 (TRA86), an increasing number of corporations chose to organize as S corporations (Nelson, 1991; Plesko, 1994). Lower individual income tax rates relative to corporate tax rates increased the tax benefits associated with S corporation election. Recent research suggests that tax rates play a key role in a firm's choice of organizational form (Gordon and MacKie-Mason, 1991; Carroll and Joulfaian, 1994; Gentry, 1994; Plesko, 1994; Ayers et al., 1996).

[3] The tax code treats the income of S corporations in a manner similar to that of partnerships. An S corporation's profits or losses are passed through to shareholders (i.e., only taxable on their personal tax returns) and thus avoid the double taxation of corporate earnings. S corporations, however, face a number of restrictions. For example, prior to 1997, S corporations could have no more than 35 shareholders, could not issue more than one class of stock, nor be part of an affiliated group, were required to be a domestic corporation, and, once S corporation status was revoked, a firm could not reelect S corporation status for 5 years. /1/ These restrictions on S corporations may inhibit the growth of firms that expect to grow by constraining their ability to raise capital through expansion of ownership and stock issues.

[4] While the choice of organizational form merits attention, the consequences of the choice on a firm's performance have been overlooked in the literature. Firms that converted from C to S status in the aftermath of the TRA86 may have reduced their taxes significantly. At the theoretical level, if changes in taxes lead to changes in the behavior of these corporations, then we should expect to observe faster growth as well. However, it is not clear whether this reduction in taxes, and the concomitant decrease in the cost of capital and increase in labor supply incentives, has led to an increase in output or improvement in firm performance, as firms electing S corporation status may have limited their ability to raise capital relative to similar C corporations. /2/

[5] In this paper, we investigate the determinants of S corporation elections with an emphasis on the role of taxes. In addition, this paper is the first to attempt to study the effect of S election on a firm's post-TRA86 sales, controlling for its pre-TRA86 attributes. We use a panel of corporate income tax returns for tax years 1985 and 1990. These years span the TRA86 which brought about significant changes in tax rates. In studying organizational choice, our data help avoid aggregation bias common to group data (MacKie- Mason and Gordon, 1991), are superior to cross-sectional data (Carroll and Joulfaian, 1994; Ayers et al., 1994) /3/, and avoid sample selectivity bias common to studies using panel data on corporations with a small number of shareholders, where the latter is defined using post-TRA86, post-behavior data on the number of shareholders (Plesko, 1994). /4/

[6] The next section briefly summarizes the growing importance of S corporations since the TRA86. The third section develops a theoretical model of firm choice of organizational form. The fourth section discusses the data sources, development of our panel, and variable construction. The fifth section presents the empirical results. The final section concludes the paper.

II. RISE IN S CORPORATIONS

[7] The changes in tax rates brought about by the TRA86 led to a significant increase in S corporation elections. Prior to TRA86, the maximum statutory tax rate for corporations was 46 percent, with 50 percent for the non-corporate sector. Under the TRA86, a firm organized as a C corporation faced a maximum tax rate of 34 percent. /5/ This firm may distribute after-tax income in the form of dividends or retain such income for future investments. Distributed dividends will be taxed as ordinary income, while retained earnings will be taxed as capital gains when realized, both at a maximum statutory tax rate of 28 percent. In contrast, S corporations are able to pass income through to the shareholders directly and avoid double taxation. The income of an S corporation is taxed only once at the individual (non-corporate) level as ordinary income.

[8] Table 1 presents tabulations of data from the Corporate Statistics of Income (SOI) to illustrate the growth in S corporations following the TRA86. The increase in both the number of S corporations and their income after the TRA86 was dramatic. The number of firms electing S corporation status rose two-fold, from 0.7 million to 1.4 million between 1984 and 1990. During this period, S corporations as a share of all corporations rose from 22 percent to 42 percent. /6/ The rise in S corporations is equally dramatic when income is used to measure their growth. S corporation income as a share of total corporate income rose from 5.4 percent to 9.3 percent between 1984 and 1990.

[9] The growth in assets and sales is even more dramatic. The share of corporate assets held by S corporations rose from 1.6 percent in 1984 to 4.1 percent in 1990. /7/ The comparable figures for sales are 6.3 percent and 16.1 percent, respectively. Whether the number, income, or assets are used to measure growth, the growth of S corporations was very dramatic after the TRA86. In fact, some researchers have pointed to the growth of S corporations as a significant source of erosion for the corporate income tax base after the TRA86, depressing corporate tax receipts during the late 1980's (for example, see Gordon and MacKie-Mason, 1991; Poterba, 1991).

[10] Table 1 also shows that the increase in S corporation elections was strongest in 1987 tapering off quickly in later years. These trends are also evident when using income or assets. The representation of S corporations within the corporate sector stabilized by the end of the decade. These trends may reflect the timing of the tax rate changes. The largest fall in the individual/corporate rate differential occurred between 1986 and 1987, while the individual rate cuts were not fully phased-in until 1988.

[11] A comparison of S corporation status by industry, also presented on Table 1, indicates that S corporation elections are more important for some industries than others. By 1990 over 32 percent of all S corporations were in the services industry. The dominance of S corporations in the services industry, and to a lesser extent wholesale and retail trade industries, probably reflects the smaller size of firms in these industries.

III. AN ANALYTICAL FRAMEWORK OF CORPORATE ORGANIZATIONAL CHOICE

[12] Entrepreneurs may choose from four forms of legal organization: partnership, proprietorship, C corporation, and S corporation. The choice between the corporate and noncorporate forms is a complex one and beyond the scope of this paper. /8/ Once a decision is made to incorporate, firms will compare the tax benefit from S corporation election to the non-tax benefits of C corporation election. Prior to the changes enacted by the Congress in 1996, S corporations were not able to reap many of the benefits available to publicly traded firms or firms with many shareholders because S corporations were limited to 35 or fewer shareholders and one class of stock. /9/ Thus, in choosing between the two forms of corporate organization, a firm would compare the expected after-tax income under each organizational form or regime. /10/ If the firm decides to become an S corporation, its after-tax return or profit is given by,

p[sub s] = Y (1 - t[sub nc]) (1)

where p is after-tax profit, Y is pre-tax profit, t[sub nc] is the

[13] In contrast, the after-tax profit of a firm electing to become a C corporation is given by,

p[sub c] = (Y + vA) (1 - t[sub c]) (2)

where,

t[sub c] - t + (1 - t) [(1 - a) t[sub cg] + at[sub d]],

v is the expected additional earnings from being a C corporation as a

[14] A firm is indifferent to its choice at the point where the tax benefits for S corporation election are equal to the non-tax benefits from C corporation status, or

Y (t[sub c] - t[sub nc] = vA (1 - t[subc]) (3)

Equation (3) assumes that tax considerations are proportional to

[15] To derive the probability of a firm choosing S corporation statuts from equation (3) we assume that there is stochastic variation in behavior across firms such that the standard deviation of the error term is proportional to the size of the firm's assets, A, and rewrite equation (3) as,

Y (t[sub c] - t[sub nc]) = vAe[super E] (1 - t[sub c]) (3')

It follows that the value of E at which the frim is indifferent to

E* [is approximately] 1n (Y/A) + 1n (t[sub c] - t[sub nc]/1 -

Thus, the cut-off value, or the S corporation choice, depends

IV. DATA AND ESTIMATION STRATEGY

[16] Panel data on corporations were developed from the 1985 and 1990 Statistics of Income (SOI) cross-sectional samples of corporate income tax returns. This period witnessed significant variations in tax rates and reflects the fully phased-in provisions of TRA86. Each cross-sectional sample consists of about 80,000 corporate tax returns prepared by the Statistics of Income (SOI) division of the Internal Revenue Service. We were able to link about 41,400 corporate tax returns across these two years. The age of each firm in the constructed panel is at least 6 years by 1990, and thus by-passes the birth and death cycle of corporations.

[17] Since our paper focuses on a firm's choice between C and S status, we only want to consider firms that either are or could become S corporations. Therefore, we exclude firms with foreign shareholders, or that conduct business in the insurance and banking industries. Also, we exclude net deficit firms as well as those with no assets. /12/ After these adjustments, 20,045 tax returns remain in our panel.

[18] These data allow us to identity a firm's choice of organizational form both before and after the TRA86. Hence, we can consider all possible transitions between C corporation and S corporation status between the two years. There are four possible outcomes: (1) C corporations in 1985 could have chosen to become S corporations by 1990, (2) C corporations in 1985 could have chosen to remain as C corporations, (3) S corporations in 1985 could have chosen to become C corporations, or (4) S corporations in 1985 could have chosen to remain as S corporations.

[19] Table 2 provides a transition matrix of the choices firms made between C and S status in our panel. Of the 20,045 firms in the panel, 15,302 (76 percent) elected C corporation status in 1990 and the remaining 4,743 (24 percent) elected S corporation status. However, of those corporations that elected S corporation status in 1990, only 1,733 (37 percent) also had elected S corporation status in 1985. In 1985, 18,229 firms were organized as C corporations. Of these, 3,010 (17 percent) elected S corporation status in 1990, while 15,219 retained their C corporation status.

[20] We use two specifications to model the transition between organizational form. First, we consider a firm's choice of organizational form in 1990 controlling for its choice in 1985. The dependent variable takes on a value of one if a firm elected S corporation status in 1990, zero otherwise. In this case, we include all firms that were eligible to be S corporations in 1990 regardless of their status in 1985. A lagged dependent variable indicating a firm's choice of organizational form in 1985 is included. In an alternative specification, we consider a firm's choice of organizational form in 1990 but exclude firms that were organized as S corporations in 1985. In this specification we restrict the sample to firms organized as C corporations in 1985. A motivation for this restriction is that income and asset statements of C and S corporations are very likely to have been constructed differently.

[21] Table 3 provides the means and standard deviations of select variables by type of organizational form elected in 1990. We begin with the ratio of income to assets in 1985, which reflects the gross profitability of the firm. Since this variable is based on a firm's pre-TRA86 attributes, it is exogenous to a firm's organizational choice in 1990. /13/ Income is defined as net income reported on a firm's corporate income tax form (Form 1120, line 28). Assets are obtained from the balance sheet information reported on Schedule L. For the panel of 20,045 corporations, the average asset size is $248 million, and the ratio of income to assets is 0.202. For the 18,229 corporations electing C status in 1985 (column 2), average assets are $270 million with an average ratio of income to assets of 0.102.

[22] The change in the grossed-up tax wedge between the corporate and non-corporate sector, the second term in equation (4), reflects the change in the advantage of S corporation election due to the TRA86. For each firm, we compute tax liability as if it were a C corporation and divide it by income to compute t /14/ Although we are able to compute each firm's corporate level tax rate, t, we do not have any information on each shareholder's individual tax rate, t[sub d], nor their tax rate on capital gains and holding period needed to compute t[sub cg]. Therefore, we set the tax rate on dividends and non-corporate income at 0.37 in 1985 and 0.28 in 1990. Similarly the capital gains tax rates are set at 0.10 and 0.14, respectively. /15/ We use aggregate data and set the dividend payout rate, a, to 0.6, which is exogenous to the choice of organizational form. The average reduction in the grossed-up tax wedge, denoted by T in Table 3, is 13.7 percent over the period. The reduction in the grossed-up tax wedge computed for the 18,179 firms organized as C corporations in 1985 is 13.1 percent.

[23] Table 3 also provides information on the industrial classification of firms in the panel. About two percent of the firms in our panel are in agriculture, one percent in mining, seven percent in construction, four percent in transportation and utilities, 15 percent in retail trade, 14 percent in wholesale trade, 27 percent in finance, and 9 percent in services.

V. RESULTS

1. Choice of Organizational Form

[24] Table 4 provides weighted probit estimates for S corporation elections in 1990. Column 1 of Table 4 provides estimates for the panel of 20,045 firms. /16/ The change in the grossed-up tax wedge from equation (4), is included to capture the effects of the tax rate changes under TRA86 on organizational form. Additional variables are also included to control for firm characteristics in 1985, prior to the changes under TRA86. These include variables such as the ratio of income to assets, organizational form in 1985, and dummy variables for major industry to control for otherwise unobservable industry-specific effects. /17/

[25] The coefficient on the ratio of income to assets is positive, consistent with theory, and is estimated at 0.16 with a standard error of 0.012. The marginal effect is 0.02 which yields an elasticity of 0.16 for the probability of electing S status with respect to a firm's return on its assets. This result implies that greater profitability encourages firms to organize as S corporations.

[26] The tax rate coefficient is positive, also consistent with theory, with an estimate of 0.20 and standard error of 0.023. This indicates that firms with the greatest increase in the corporate- noncorporate tax wedge between 1985 and 1990 are more likely to have elected S corporation status, where the tax savings are greatest for profitable firms. The marginal effect is 0.03, with an implied elasticity of 0.20 for the probability of S status election. This low elasticity coefficient should not be surprising because larger C corporations may find that the cost imposed by the restrictions of S corporation status (or the benefits of C status), such as the number of shareholders and access to capital, outweigh its tax advantages and simply retain their C corporation status. /18/

[27] Column 2 of Table 4 replicates the profit estimates of column 1 but limited to the 18,179 firms organized as C corporations in 1985. Because all the S corporations in this subpanel had elected the C status in 1985, they represent firms that switched their status over the two years. The estimated coefficients, as well as a computed marginal effects and elasticity coefficients, are very similar to those observed in column 1, except for the industry coefficients.

[28] Both estimated coefficients of the tax wedge reported in columns 1 and 2 of Table 4 imply an elasticity of about 0.2. This would suggest that, in the absence of the rate changes brought about by TRA86, the number of profitable C corporations would have been 20 percent higher. This would roughly translate into an additional 3,000 corporations choosing to retain C corporation status in our sample, or about 360,000 in the population.

2. Organizational form and firm performance

[29] Moving to the effect of S elections on firm performance, Table 5 presents ordinary least squares estimates for the growth rate of firm sales between 1985 and 1990. /19/ The dependent variable is defined as the difference of the natural logarithm of sales over the two periods. We use the amount of sales as the dependent variable as a proxy for firm output which is not observed in the data. The explanatory variables consist of pre-TRA86 attributes and include sales in 1985, industrial classification to control for industry specific effects, and indicators of whether a firm was an S corporation in 1985 or had switched to S corporation status by 1990.

[30] Controlling for size and industry, weighted regression results in column (1) show that the sales of S corporations grew by 32 percent and that C firms that had switched to the S status between 1985 and 1990 grew 23 percent faster than the remaining C corporations. Similar results are obtained in column (2) which is restricted to firms organized as S corporations in 1985. This specification indicates that the sales of firms that switched to the S status in the aftermath of TRA86 grew 18 percent faster than the remaining C corporations over the period. /20/ While these estimated response parameters seem large, they are well below the growth pattern observed in Table 1 which shows that S corporations' share of receipts more than tripled between 1985 and 1990 while their share of the number of corporations only doubled.

[31] In an alternative specification, we limit the sample to firms that are of comparable size. Specifically, we exclude C corporations with assets in excess of the largest S corporation in 1990 (slightly over $1 billion). We do this for two reasons: (1) firms with tens and hundreds of billions in assets may find it very difficult to convert, and (2) very large firms may grow at slow rates, which may bias our estimates. Both profit and regression estimates were little affected using the reduced sample.

VI. CONCLUSION

[32] This paper investigates the determinants of the corporate choice of organizational form and the effect of this choice on firm performance. Using a panel of corporation income tax returns before and after the Tax Reform Act of 1986, we evaluate the role of taxes on the election of S corporation status.

[33] Our results show that higher corporate-noncorporate tax rate differentials increase the likelihood that a firm will convert from C to S corporation status, where the tax savings are the greatest for profitable firms. Such effect, though significant, is likely to be small for established firms, which tend to be large and require greater access to capital. This is consistent with the observed pattern of low S elections by older firms, but high S elections by newer firms. This may indicate that non-tax factors (e.g., access to capital) may mitigate the distortionary effects of the differential tax treatment and the consequences for government revenue. This, however, may need to be directly verified by studying the longitudinal pattern of organizational choice of S corporations.

[34] Our results also indicate that firms organized as S corporations tend to grow faster than C corporations. For the average firm in our sample, the sales of C corporations that switched to S status after TRA86 grew about 20 percent faster than the average corporation during the 1985 and 1990 period.

FOOTNOTES

We thank Bill Gentry, Lowell Dworin, Michael McDonald, Laura Kalambokidis, Susan Nelson, Tom Neubig, George Plesko, and Karl Scholz for their helpful comments and suggestions.

/1/ The Small Business Job Protection Act of 1996 included a number of provisions affecting S corporations. For example, this Act increased the limit on the maximum number of shareholders for S corporations from 35 to 75, to some extent liberalized the rules governing who can hold shares in an S corporation (e.g., certain trusts and tax-exempt entities) and under what circumstances S corporations can own part or all of another corporation, and extended the option of organizing as an S corporation to other types of businesses (e.g., certain financial institutions).

/2/ See Gravelle and Kotlikoff (1989). For a contrarian view see Gordon and MacKie-Mason (1994).

/3/ These studies focus on small corporations.

/4/ The sample selection problem may bias the results because C corporations with large numbers of shareholders that convert to S corporations by reducing the number of shareholders are included in panels that define the number of shareholders using post-TRA86, post- behavior data.

/5/ The Omnibus Budget Reconciliation Act of 1993 raised the maximum corporate tax rate for C corporations to 35 percent.

/6/ Although not reported on Table 1, the fraction of new incorporations electing S status rose from 30 to 55 percent.

/7/ Corporations comprise a small share of the corporate sector when income or assets is used, rather than returns, because S corporations tend to be smaller.

/8/ Limited liability laws are likely to play a dominant role in the choice between the two forms of organization. Papers that address diverse aspects of the effect of the tax treatment of the corporate and noncorporate sectors include Gentry (1991), Guenther (1992), Jones and Taggart (1984), and Terando and Omer (1993).

/9/ The Small Business Job Protection Act of 1996 limited S corporations to 75 or fewer shareholders.

/10/ The theoretical model provided below draws on MacKie-Mason and Gordon (1991). Also see Scholes and Wolfson (1992, chapter 4).

/11/ As described in greater detail below, we adjust the tax rate of capital gains relizations to account for dererrals.

/12/ Net deficit firms may have an incentive to elect S corporation status in the face of relatively higher individual tax rates to take advantage of the pass-through of losses. To the extent that the corporate tax rate is not observed in the presence of carry- forwards, even though it is zero at present, we are unable to address the role of taxes for these firms. Also note that information on the number of shareholders is not available in 1985; information in the 1990 data is limited to 35 or fewer shareholders.

/13/ Note that using 1990 information does not change the qualitative results.

/14/ One difference between S and C corporations is that S corporations do not report a deduction for net operating losses (NOLs) on their corporate tax returns nor certain tax credits. We account for these differences by computing tax rates for all firms before NOLs and credits.

/15/ All the tax rates are computed for individuals with AGI in excess of $50,000 using SOI individual data. We assume that deferrals reduce the tax rate on capital gains by one-half.

/16/ S corporations tend to be under sampled in the cross- sectional files used to construct our panel, as a comparison of Table 1 and 2 would suggest. To correct for this under representation, the probit and ordinary least squares equations presented below are estimated using 1985 weights. Qualitatively similar results are obtained using the greater of the 1985 and 1990 weights.

/17/ The manufacturing industry dummy is excluded from the model to allow estimation.

/18/ This is consistent with findings from the 1990 SOI file which show that only one third of the firms incorporated in 1985 or earlier are organized as S corporations, while 55 percent of firms incorporated in later years made this election. (Additional information is available upon request.) In addition to the number of shareholders, recapture taxes, which we are unable to model, resulting from C to S conversions are more binding for older firms such as those in the panel. Note that in addition to the omitted variable (benefit of C corporation) bias, the estimate may suffer from aggregation bias since we do not observe the individual shareholder's tax rate.

/19/ As an alternative to sales, one may compare income or asset changes over time. Because income and asset statements of C and S corporations may be constructed differently, however, such comparisons may not be appropriate.

/20/ Similar results are obtained using a log-log specification, in place of growth rates, except that the adjusted R square more than doubles in value.

END OF FOOTNOTES

REFERENCES

Ayers, Benjamin C., C. Bryan Cloyd, and John R. Robinson (1996).

Carroll, Robert and David Joulfaian (1994). "Do Taxes Affect

Gentry, William M. (1991). "Taxes and Organizational Form: The Rise

_____ (1994). "Taxes, Financial Decision and Organizational Form:

Gordon, Roger G. and Jeffrey MacKie-Mason (1991). "Effects of the tax

Gravelle, Jane G. and Larry J. Kotlikoff (1989). "The Incidence and

Guenther, D.A. (1992). "Taxes and Organizational Form: A Comparison

Jones, E. Philip and Robert A. Taggart (1984). "Taxes and Ownership

Internal Revenue Service, Statistics of Income -- Corporation Income

MacKie-Mason, Jeffrey and Roger Gordon (1991). "Taxes and the Choice

______ (1994). "Tax Distortions to the Choice of Organizational

Nelson, Susan (1991). "S Corporations Since the Tax Reform Act of

Plesko, George A. (1994). "Entity Choice: Financial Characteristics

______. (1994). "The Role of Taxes in Organizational Choice: S

Poterba, James M. (1991). "Why Didn't the Tax Reform Act of 1986

Scholes, Myron S. and Mark K. Wolfson, Taxes and Business Strategy: A

Terando, William D. and Thomas C. Omer, (1993). "Corporate

* * * * *

                               TABLE 1

                 THE RISE IN S CORPORATION ELECTIONS

NUMBER OF S CORPORATIONS (IN THOUSANDS)

  All S Corporations                        701        725        826

S CORPORATIONS AS SHARE OF ALL CORPORATIONS

  All S Corporations                      22.1%      22.1%      24.1%

ANNUAL GROWTH IN S CORPORATIONS

  All S Corporations                                  3.3%      14.0%

S CORPORATION INDUSTRY SHARES

  Agriculture                              4.8%       4.5%       4.2%

S CORPORATION INCOME AS SHARE OF ALL CORPORATE INCOME

  Total Receipts                           6.3%       5.1%       5.6%

                         [table 1 continued]

_____________________________________________________________________

NUMBER OF S CORPORATIONS

  All S Corporations                1,128    1,257    1,423     1,575

S CORPORATIONS AS SHARE OF ALL CORPORATIONS

  All S Corporations                31.2%    35.3%    39.2%     42.4%

ANNUAL GROWTH IN S CORPORATIONS

  All S Corporations                36.5%    11.5%    13.2%      10.7

S CORPORATION INDUSTRY SHARES

  Agriculture                        4.2%     4.0%     3.7%      3.5%

S CORPORATION INCOME AS SHARE OF ALL CORPORATE INCOME

  Total Receipts                   10.1%     12.3%    13.4%     16.1%

     SOURCE: Statistic of Income, Corporate Income Tax Returns,

                              * * * * *

                               TABLE 2

         TRANSITIONS BETWEEN CORPORATE ORGANIZATIONAL FORMS

                                1990

                         C Corporation    S Corporation     Total

        C Corporation        15,219          3,010          18,229

        Total                15,302          4,743          20,045

                              * * * * *

                               TABLE 3

      SAMPLE MEANS AND STANDARD DEVIATIONS OF SELECT VARIABLES

                                                   Mean

d ln T                                  -0.137             -0.131

Assets ($1,000)                        247,900            270,030

ln Assets                               17.462             17.531

Income/assets                            0.202              0.102

ln Income/assets                        -2.911             -3.009

Agriculture                              0.024              0.022

Mining                                   0.011              0.010

Construction                             0.075              0.073

Utilities                                0.039              0.039

Wholesale Trade                          0.136              0.031

Retail Trade                             0.149              0.148

Finance, Insurance, and                  0.265              0.284

Services                                 0.090              0.083

Number of Corporations                  20,045             18,229

S Corporations in 1990                   4,743              3,010

T = [t[subc] - t[sub nc]]/[1-t[subc]]

                         FOOTNOTE TO TABLE 3

     /*/ In 1985.

                      END OF FOOTNOTE TO TABLE

____________________________________________________________________

                              * * * * *

                               TABLE 4

             PROFIT ESTIMATES FOR S CORPORATION ELECTION

Variable                                  (1)          (2)

Constant                                -1.690 /*/  -2.549 /*/

ln (Income / Assets)                     0.157 /*/   0.143 /*/

d ln T                                   0.196 /*/   0.187 /*/

Agriculture                              0.256 /*/  -0.262

Mining                                   0.284 /*/   0.342

Construction                             0.625 /*/   1.708 /*/

Utilities                                0.829 /*/   1.607 /*/

Wholesale Trade                          0.729 /*/   1.554 /*/

Retail Trade                             0.316 /*/   1.091 /*/

Finance, Insurance, and                  0.338 /*/   1.085 /*/

Services                                 1.064 /*/   1.948 /*/

S Corp in 1985                           3.008 /*/    --

Log Likelihood                          -5,264      -5,590

Pseudo-R(super 2)                         0.61        0.53

Number of Observations                  20,045      18,229

Positive observations (S Corporations)   4,743       3,010

     T = [t[sub c] - t[sub nc]]/[1-t[sub c]]

                         FOOTNOTE TO TABLE 4

     /*/ Estimates are statistically significant at the 5 percent

                      END OF FOOTNOTE TO TABLE

                              * * * * *

                               TABLE 5

             REGRESSION ESTIMATES OF THE GROWTH OF SALES

____________________________________________________________________

Constant                              6.682 /*/        6.146 /*/

ln Sales in 1985                     -0.970 /*/       -1.004 /*/

(ln Sales in 1985)(super 2)           0.034 /*/        0.038 /*/

Agriculture                          -1.423 /*/       -1.618 /*/

Mining                               -1.013 /*/       -3.778 /*/

Construction                         -0.069            0.220 /*/

Utilities                            -0.409 /*/       -0.534 /*/

Wholesale Trade                      -0.458 /*/      -0.673 /*/

Retail Trade                         -0.407 /*/       0.129

Finance, Insurance, and              -0.453 /*/      -0.093

Services                              0.137           0.172 /*/

S Corporation in 1985                 0.324 /*/         --

SwitcHed from C to S                  0.227 /*/        0.178 /*/

R(super 2)                            0.195            0.203

Number of Observations               20,045           18,229

                         FOOTNOTE TO TABLE 5

     /*/ Estimates are statistically significant at the 5 percent

                      END OF FOOTNOTE TO TABLE