Who Benefits from the Tax Advantages of Organizational Form Choice?

National Tax Journal, December 2015, 68 (4), 975–998 http://dx.doi.org/10.17310/ntj.2015.4.04

WHO BENEFITS FROM THE TAX ADVANTAGES OF ORGANIZATIONAL FORM CHOICE?

Michael P. Donohoe, Petro Lisowsky, and Michael A. Mayberry

We investigate whether and the extent to which four stakeholders of a firm — customers, suppliers, employees, and owners — benefit from the savings under Subchapter S relative to Subchapter C. Using a balanced sample of U.S. commercial banks from 2001–2005, we find that S corporation banks pay 2 percent more in wages and 38 percent more in dividends than C corporation banks, but pay lower interest rates on deposits and charge similar rates on loans. Further, after the Jobs and Growth Tax Relief and Reconciliation Act of 2003 attenuated the relative tax advantage of Subchapter S, we find that S corporation banks increased wages and dividends to a lesser extent than C corporation banks. These results suggest that employees and owners, but not suppliers or customers, benefit from the tax advantages of the Subchapter S organizational form. Overall, our study provides economic insight into which parties ultimately bear — and do not bear — the incidence of corporate taxation.

Keywords: tax incidence, organizational form, S corporation, C corporation, banks JEL Codes: G21, H25, M40

I. INTRODUCTION usinesses operate under various incorporated and unincorporated organizational Bforms, each with their own unique tax and non-tax features (Scholes et al., 2014). The corporate form provides centralized management, perpetual life, transferable ownership, and limited personal liability to its owners. However, income can be subject to taxation at the owner and entity level. In contrast, while partnerships and sole propri-

Michael P. Donohoe: University of Illinois at Urbana-Champaign, Champaign, IL, USA (mdonohoe@ illinois.edu) Petro Lisowsky: University of Illinois at Urbana-Champaign, Champaign, IL, USA; Massachusetts Institute of Technology, Cambridge, MA, USA; and Norwegian Center for Taxation, Bergen, Norway (lisowsky@ illinois.edu) Michael A. Mayberry: University of Florida, Gainesville, FL, USA ([email protected]) 976 National Tax Journal etorships do not offer such non-tax benefits of incorporation, they incur only one layer of income tax — at the owner level. Prior research finds that income taxes, due to their material cost, are an important consideration in organizational form choice (Hanlon and Heitzman, 2010). Nevertheless, studies identifying whether various stakeholders of a firm reap any tax-related benefits of organizational form are largely absent from the literature (Auerbach, 2006). We fill this void by investigating whether and to what extent four parties — customers, suppliers, employees, and owners — benefit from the relative tax advantages of organizational form choice. To isolate the tax advantages of organizational form, we consider two corporate forms that differ primarily on their treatment under the income tax: Subchapter C corporations and Subchapter S corporations. Subchapter C of the Internal Revenue Code (IRC) lev- ies an entity-level tax on corporate profits, which upon distribution (e.g., as dividends) are taxed again under the individual income tax. This structure is often referred to as double taxation. Under Subchapter S, however, corporations meeting specific criteria can elect to avoid the entity-level tax and instead “pass through” income, losses, deduc- tions, and credits directly to shareholders, who are only subject to individual income taxes (Internal Revenue Service (IRS), 2013a).1 This single-layer tax advantage can be economically large. For instance, Donohoe, Lisowsky, and Mayberry (2015) estimate that 740 C corporation banks converting to S status from 1997–2010 realized tax savings of about 9.6 percent of net income in the first year alone. As a result, S corporations are a popular organizational form, comprising 44 percent of all business tax returns filed during 2012 (IRS, 2013b). Apart from income taxes, Subchapter C and S corporations are legally the same type of entity. Subchapter S simply combines the legal benefits of incorporation with tax benefits similar to that of partnerships or sole proprietorships. Therefore, comparing S and C corporations offers a unique and powerful setting in which to identify the tax advantages of organizational form choice. We use the setting of depository institutions (i.e., commercial banks) to investigate our research question because it offers several empirical advantages. First, both C and S corporation banks file Call Reports with the Federal Deposit Insurance Corporation (FDIC) that specify organizational form. Other well-known databases, such as Compus- tat, largely contain only publicly-traded C corporations. Thus, identifying privately-held C and S corporations allows us to consider a broad portion of the economy in evaluat- ing the consequences of organizational form choice.2 Second, Call Reports provide a rich array of financial information, such as wages paid to employees, dividends paid to owners, and interest on deposits and loans. Presently, Compustat does not provide such granularity in the disclosure of income and expense items. This information allows us to identify which of the four different stakeholders benefit from the tax advantages of organizational form. Finally, the banking industry is relatively homogenous in terms

1 The S corporation form has significant restrictions, including limitations on stock classes as well as the number and types of shareholders. See Section II for the relevant criteria. 2 In addition, commercial banks cannot select an organizational form other than C or S corporations (e.g., a partnership). This feature allows us to examine all available organizational forms for commercial banks. Who Benefits from the Tax Advantages of Organizational Form Choice? 977 of product offerings (loans), operations, accounting, and balance sheet composition (Beaver, McAnally, and Stinson, 1997). Focusing on a relatively homogenous sample mitigates cross-industry and non-tax differences as alternative explanations of organizational form choice, which has been a challenge for prior research on corporate tax incidence (Sahni and Mathew, 1976; Auerbach, 2006).3 Using a balanced sample of U.S.-headquartered commercial banks from 2001–2005, we find that tax-advantaged S corporation banks pay 2 percent more in wages and 38 percent more dividends than tax-disadvantaged C corporation banks. However, C and S corporation banks offer similar loan interest rates, while S banks pay lower interest rates on deposits. These results are consistent with employees and owners, but not suppliers or customers, benefiting from the tax advantages of S corporation status. Further, because employees and owners of C corporations are paid less than those of S corporations merely due to the different tax effects of organizational form, our results suggest that these parties bear at least some of the corporate tax burden. In further tests using a difference-in-differences design, we find that S corporation banks increased wages and dividends to a lesser extent than C corporation banks after the Jobs and Growth Tax Relief and Reconciliation Act of 2003 (JGTRRA) reduced the relative tax advantage of Subchapter S. In particular, JGTRRA lessened the “sting” of double taxation by reducing the dividend tax rate from 38.6 to 15 percent, while providing only slight tax benefits to S corporation shareholders by reducing ordinary personal income tax rates (on which their share of S corporation profits are based) from 38.6 to 35 percent. This relatively large exogenous reduction in dividend tax rates relative to ordinary personal tax rates provides a powerful setting to validate that our findings are attributable to tax — rather than unobservable non-tax — factors between S and C corporations. We estimate that after JGTRRA, C banks increased wages and dividends by approximately 15 and 2 percent, respectively, while S banks increased wages by only 14 percent and did not significantly increase dividends. These results have two implications. First, they suggest that the reduction in shareholder tax rates not only directly benefited owners, but also employees, particularly for tax-disadvantaged C corporations. Second, the JGTRRA tests confirm our prior inference that employees and owners bear some portion of the incidence of corporate-level taxes. That is, when JGTRRA reduced the relative tax disadvantage of C over S status, C corporations nar- rowed the gap on wage and dividend payments compared to S corporations. Our study contributes to research on taxes and organizational form (e.g., Shevlin, 1987; Petroni and Shackelford, 1995; Ayers, Cloyd, and Robinson, 1996; Donohoe, Lisowsky, and Mayberry, 2015) by providing economic insight into which parties ultimately bear — and do not bear — the burden of corporate taxation. It also provides new insights into how the tax attributes of corporations can have a real effect on employees. Finally,

3 Traditionally, tax incidence is based on capital reallocation changing pre-tax returns in an open-economy general equilibrium model (Clausing, 2013). However, consistent with recent research (Carroll, 2009; Arulampalam, Devereux, and Maffini, 2010; Fuest, Peichl, and Siegloch, 2013; Liu and Altshuler, 2013), we define tax incidence more broadly as both capital reallocation and the effects of taxation on rent sharing. 978 National Tax Journal we contribute to the literature on depository institutions (Ryan, 2007) by identifying tax factors that appear to affect payout policies to bank employees and owners, but interestingly, not depositors or borrowers. We proceed as follows. Section II provides background information and develops our hypotheses. Our empirical methodology and main results are described in Sections III and IV, respectively. Section V reports robustness tests, and Section VI concludes.

II. BACKGROUND and HYPOTHESIS DEVELOPMENT

A. Organizational Form The choice of organizational form affects numerous facets of business, including taxation, access to capital, and liability exposure (Ayers, Cloyd, and Robinson, 1996). For example, MacKie-Mason and Gordon (1997) show that profitable firms are reluc- tant to adopt the corporate form due to the tax costs of operating as a C corporation. More recently, Allen and Raghavan (2012) find that some loss firms organize as C corporations rather than more tax-advantaged forms due to the operating preferences of private equity investors. Extant research also suggests that income taxes explain the use of general partnerships (Guenther, 1992), limited partnerships (Shevlin, 1987), master limited partnerships (Terando and Omer, 1993; Gentry, 1994; Shelley, Omer, and Atwood, 1998), and real estate investment trusts (Goolsbee and Maydew, 2002). Although taxation plays a role in organizational form choice, its relative importance is unclear. For instance, Goolsbee (1998) finds that the effect of taxes on organizational form choice was negligible during 1900–1939, while other research suggests non-tax factors (e.g., financial reporting benefits, business risk, firm age, and firm size) affect organizational form choices to a greater extent than tax factors (Beatty, Berger, and Magliolo, 1995; Ayers, Cloyd, and Robinson, 1996). Evidence on the tax consequences of organizational form is also rather limited. For example, Erickson and Wang (2007) conclude that mergers with S corporations elicit a higher price premium relative to mergers with C corporations, and Erickson and Wang (2000) find that tax elections that generate future tax savings for acquirers lead to higher purchase prices. This evidence suggests that buyers and sellers of a firm share the tax advantages of Subchapter S. In this study, we consider two incorporated organizational forms: Subchapter C and Subchapter S corporations. S corporations and C corporations differ almost entirely on tax features. Historically, in the United States corporate profits are subject to two levels of taxation under Subchapter C. First, C corporations incur an entity-level tax at corporate rates as they earn income. Second, shareholders incur an individual-level tax at shareholder rates upon the receipt of dividends. However, in 1958, Congress amended the IRC to include Subchapter S, which allowed corporations meeting specific criteria to elect out of Subchapter C’s entity-level income tax. In lieu of an entity-level tax, an S corporation’s shareholders pay tax at individual ordinary rates on their percentage share of earned (rather than distributed) corporate income. Thus, S corporations pay a single layer of tax at the shareholder level rather than the double layer of tax Who Benefits from the Tax Advantages of Organizational Form Choice? 979 faced by C corporations. As predominately a tax election, Subchapter S offers the traditional benefits of incorporation combined with the tax benefits of a pass-through entity. Despite these benefits, S corporations face restrictions and disadvantages. For example, only a limited number and specific types of shareholders can own S corporations. To qualify, an S corporation must: (1) be incorporated domestically; (2) have only individual and certain trust/estate shareholders (i.e., partnerships, corporations, and non-resident alien shareholders are prohibited); (3) have a maximum of 100 shareholders; and (4) have only a single class of stock. The S corporation form also provides for fewer tax deferral opportunities as shareholders are immediately taxed on their share of income, regardless of whether income is actually distributed.

B. Hypothesis Development Table 1, adapted from Donohoe, Lisowsky, and Mayberry (2015), provides a stylized example of the tax savings of S corporations relative to C corporations for the corporation and its owners, both before and after JGTRRA reduced both the dividend tax rate (from 38.6 to 15 percent) and ordinary personal tax rates (from 38.6 to 35 percent). In the example, we use the average dividend payout ratio for our sample of 39 percent. We note three important features. First, the double taxation of C corporations provides S corporations with a relative tax advantage in that income is only taxed once at the owner level. Although the income allocated to S corporation shareholders is immediately taxed at ordinary personal rates, which is similar to the corporate rate, S corporations do not pay entity-level taxes and their owners do not pay taxes on dividends received. Second, JGTRRA reduced the overall tax burden for both C and S corporations, with after-tax income increasing by 18.43 and 5.86 percent, respectively. The reduction in dividend tax rates benefited C corporation owners by reducing the level of double taxation, while the reduction in personal tax rates provided a benefit, albeit smaller, to S corporation owners. Third, despite an increase in absolute tax benefits for both S and C corporations after JGTRRA, the relative tax advantage of S over C corporations declined from 22.93 to 9.89 percent.4 Auerbach (2006) reviews the economic theory and evidence about which parties bear the corporate income tax burden. Early research suggests that firm owners bear a large portion of the burden, while later research (Harberger, 1962) concludes that owners bear the burden in proportion to the overall level of corporate capital within an economy, with the non-corporate sector bearing the residual burden. Harberger (1962) also finds that customers, suppliers, and employees bear no burden. More recently, Clausing

4 Before 2003 and after 2012, the top statutory tax rate for the highest ordinary income bracket was higher for individual taxpayers than corporate taxpayers. Assuming a C corporation does not pay dividends, this would make the income tax burden of a S corporation greater than a C corporation. However, in our sample, 73 percent (95 percent) of C (S) corporation banks pay dividends, suggesting that the combined entity and shareholder federal income tax burden of a S corporation is likely lower than that for a C corporation in most cases. 980 National Tax Journal

Table 1 Stylized Example of the Tax Benefits of S Corporations Relative to C Corporations

Panel A: Prior to JGTRAA of 2003 C Corporation S Corporation Net Benefit (%) Pre-tax income ($) 100.00 100.00 Corporate tax rate (%) 35.00 — Individual tax rate (%) — 38.60 Income tax ($) 35.00 38.60

Shareholder distribution1 ($) 39.00 39.00 Dividend tax rate (%) 38.60 — Dividend tax ($) 15.05 —

Total tax ($) 50.05 38.60

After-tax income ($) 49.95 61.40 22.93

Panel B: After JGTRAA of 2003 C Corporation S Corporation Net Benefit (%) Pre-tax income ($) 100.00 100.00 Corporate tax rate (%) 35.00 — Individual tax rate (%) — 35.00 Income tax ($) 35.00 35.00

Shareholder distribution1 ($) 39.00 39.00 Dividend tax rate (%) 15.00 — Dividend tax ($) 5.85 —

Total tax ($) 40.85 35.00

After-tax income ($) 59.15 65.00 9.89

Change in after-tax income (%) 18.43 5.86 1 Consistent with the behavior of the banks in our sample, we assume that 39 percent of pre-tax income is distributed as dividends. Source: This figure is adapted from Donohoe, Lisowsky, and Mayberry (2015). Who Benefits from the Tax Advantages of Organizational Form Choice? 981

(2013) and Gravelle (2013) find empirical evidence in support of Harberger’s (1962) contention that capital owners exclusively bear the burden of corporate taxation. How- ever, Liu and Altshuler (2013) outline the possibility that non-owner stakeholders bear some of the burden. Specifically, they note that if capital moves between jurisdictions to avoid taxation, other less mobile parties might bear some portion of the overall tax. Consistent with other recent research (Arulampalam, Devereux, and Maffini, 2012; aus dem Moore and Kasten, 2009), Liu and Altshuler (2013) conclude that employees bear a majority of the corporate tax burden. Aside from owners and employees, other stakeholders may bear a portion of corporate tax incidence. For instance, Goolsbee (1998) finds that some suppliers reap the benefits of lower taxes via higher prices, and Sebold (1979) estimates that 69 percent of changes in corporate taxes are passed on to customers. Finally, Ablett and Hart (2006, p. 47) conclude that, “despite [all] this work [on corporate tax incidence], unambiguous incidence conclusions remain elusive.” Thus, whether and the extent to which different stakeholders bear the corporate tax burden are open empirical questions. While we are interested in which stakeholders benefit from lower corporate taxes, such evidence also sheds light on who bears the burden of corporate taxes. Ex ante, it is not clear who will benefit from lower corporate taxes because competing arguments suggest several parties will reap some benefits. For instance, along the lines of Sebold (1979), if customers benefit from the tax-advantaged S corporation status, we state the following hypotheses (in alternative form):

H1a: S corporations charge lower interest rates on loans than C corporations. H1b: Interest rates on loans will decrease to a lesser extent for S corporations than for C corporations after JGTRRA.

Depositors could also benefit from the tax advantages of S status. Donohoe, Lisowsky, and Mayberry (2015) find a 4.5 percent increase in the market share of deposits among banks that convert from a C to an S corporation. They attribute this increase in deposits to increased advertising and the opening of new bank branches. Although deposits appear to increase after conversion to S corporation status, their findings leave open whether S banks, as one strategy, pass on a portion of the tax advantages to depositors through increased interest rates. To the extent that depositors benefit from S corporation status, we state the following hypotheses (in alternative form):

H2a: S corporations pay higher interest rates on deposits than C corporations. H2b: Interest rates on deposits will increase to a lesser extent for S corporations than for C corporations after JGTRRA.

Historically, banks were ineligible for Subchapter S status. However, the Small Busi- ness Job Protection Act of 1996 amended the IRC to allow banks the option of electing S status (but no other status) as of 1997, provided that other criteria (e.g., number of 982 National Tax Journal shareholders) are satisfied. One goal of this deregulation was to enhance small business lending and increase the take-home pay of workers. Thus, a stated intention of the Act was for customers and employees of banks to reap the tax benefits of Subchapter S. Moreover, Hein, Koch, and MacDonald (2005) suggest smaller commercial banks, many of which are S corporations, view attracting and retaining qualified employees as a competitive advantage. They argue that smaller banks can provide loan officers with greater flexibility and compensation compared to larger banks through relationship- building. The idea is that high quality loan officers can attract new bank customers. Thus, both employees and customers can benefit from the S corporation form. To the extent that employees benefit from S corporation status, we state the following hypotheses (in alternative form):

H3a: S corporations pay greater employee wages than C corporations. H3b: Employee wages will increase to a lesser extent for S corporations than for C corporations after JGTRRA.

Finally, based on the strict capital requirements imposed on banks by regulators, muted capital mobility suggests that owners bear the burden and, as a result, should benefit from the tax-advantaged S corporation organizational form. Insofar as owners benefit from S corporation status, we state the following hypotheses (in alternative form):

H4a: S corporations pay greater dividends to shareholders than C corporations. H4b: Dividend payments to shareholders will increase to a lesser extent for S corporations than for C corporations after JGTRRA.

III. EMPIRICAL METHODOLOGY

A. Research Design We first test H1a, H2a, H3a, and H4a using a pooled cross-sectional regression. We then test H1b, H2b, H3b, and H4b (i.e., the JGTRRA effects) using difference-in- differences (DID) (Bertrand and Mullainathan, 2003; Campbell, et al., 2013; Donohoe, 2015). A DID design is well-suited for our research question as it removes time-invariant differences between S and C corporations as well as common trends affecting both types of banks (Roberts and Whited, 2013). In particular, we estimate the following OLS models, adjusting standard errors for clustering at the bank level:

, (1) OutcomeS=+ββok1 Bank ++βεCtrl ∑Yeart + . (2) OutcomeS=+ββok12Bank ++ββPost 3SBank ×+Post βεCtrl ++∑Yeart

The dependent variable (Outcome) is one of four measures representing each of the four stakeholders of interest. First, to capture the tax effects of organizational form on Who Benefits from the Tax Advantages of Organizational Form Choice? 983 bank customers (H1), we define the interest rate charged on loans (LoanRate) as the ratio of interest income received on loans to year-end loan balance. Second, to consider the effects on bank suppliers (H2), we define the interest rate paid on deposits (DepRate) as interest expense paid on deposits divided by year-end deposits balance. Third, to reflect the effects on employees (H3), we define the average wage paid to employees (Wage) as the natural log of the ratio of total wages to the number of full-time equivalent employees. Finally, to capture the effect of taxes on bank owners (H4), we use the ratio of dividends to the absolute value of pre-tax income (Div).5 SBank is an indicator variable that equals one for Subchapter S banks and zero for

Subchapter C banks. The coefficient for SBank (b1) in (1) captures the average difference in the outcome variable between S and C banks over the entire sample period, while the same coefficient in (2) captures the difference in outcome variables between S and C banks before JGTRRA. Post in (2) equals one for observations in the post- JGTRRA period (i.e., after 2003) and zero otherwise, and reflects the average change in the outcome variable among all banks after JGTRRA. Finally, the coefficient for

SBank×Post (b3) in (2) is the DID estimator, which captures the effect of JGTRRA on the outcome variable for S banks relative to C banks. Ctrl is a vector of control variables that influence bank behavior. We include the log of the population of the state in which a bank is headquartered (Pop), as well as the change in state-level gross domestic product (GDP). To accommodate differences in bargaining power and resources (Berger and Bouwman, 2013), we include the log of bank assets (Size). Following prior studies (Blankespoor, et al., 2013; Jin, et al., 2013), we include three measures of bank performance: (1) lagged ratio of pre-tax income to assets (LagRoa); (2) ratio of non-performing loans to net loans (NPL); and (3) percentage change in loans in the current year (LoanGrow). We also control for financial health and cumulative performance with Lev, the ratio of non-deposit liabilities to total assets, and Cap, a bank’s Tier 1 capital ratio. Finally, year fixed-effects (Year) control for time-period specific shocks unrelated to JGTRRA.6

B. Sample Selection We obtain a sample of commercial banks headquartered in the United States from the Call Reports assembled by the Federal Reserve Bank of Chicago. First, following prior studies (e.g., Zhang, Farrel, and Brown, 2008), we define 2001–2002 as the

5 S corporations typically distribute sufficient cash to shareholders to cover the estimated tax payments that arise due to the pass-through nature of their income. To verify that our results are not influenced by this issue, we adjust Div by the estimated taxes paid on pass-through income. Specifically, we subtract the product of the top ordinary marginal tax rate on personal income and pre-tax income from dividend payments. The top tax rate was 39.6 percent (35 percent) before (after) JGTRRA. Our results persist after this adjustment, suggesting that this additional motivation for S corporations to pay dividends does not drive our findings. Our results are also robust to estimating (1) and (2) usingTobit regression. 6 We measure return on assets at t−1 because interest rates on loans, deposits, and wages are mechanically related to return on assets at t. Also, because S banks are not permitted to include deferred tax assets in Tier 1 capital ratios, we adjust the Tier 1 capital ratios of C banks for any deferred tax assets. 984 National Tax Journal pre-JGTRRA period and 2004–2005 as the post-JGTRRA period (34,285 bank-year observations).7 We select this time period as it contains two years before and after the enactment of JGTRRA in 2003. This approach allows for sufficient power to detect structural breaks, but is sufficiently narrow to mitigate threats to internal validity (e.g., Donohoe and McGill, 2011). We omit 2003 as research suggests that firm-level responses to JGTRRA occurred with a delay (Zhang, Farrel, and Brown, 2008; Campbell et al., 2013). We then restrict the sample to firms that report no foreign or public ownership (28,837 observations).8 We also require banks to have the necessary data to estimate (1) and (2) (26,302 observations). We then drop bank failures and conversions/revoca- tions to S status (leaving 23,411 observations). Finally, we require a balanced sample (21,204 observations) of banks that operate in a state with at least one other S bank (20,096 observations) to mitigate the likelihood that sample composition changes or local banking regulations, respectively, influence our results (Luna and Murray, 2010).9 The primary sample consists of 5,024 unique banks, 27.7 percent of which operate as S corporations. We winsorize all continuous variables at the 1st and 99th percentiles.

IV. MAIN RESULTS

A. Descriptive Statistics Table 2 reports descriptive statistics. The average interest rate on loans (LoanRate) is 7.3 percent, while the average rate on deposits (DepRate) is 2.3 percent. Reported values for control variables are consistent with prior research (Donohoe, Lisowsky, and Mayberry, 2015). In untabulated mean tests of differences, we find that relative to C banks, S banks charge slightly higher interest rates on loans (7.4 percent versus 7.3 percent, p<0.01) and pay slightly lower interest rates on deposits (2.3 percent versus 2.2 percent, p<0.01).10 Both S and C banks pay employees an average wage (Wage) of $45,900 (p>0.40). These tests suggest the tax advantages of Subchapter S do not accrue to customers, suppliers, or employees. However, S banks pay out 69.9 percent of their

7 The American Jobs Creation Act of 2004 increased the number of eligible S corporation shareholders from 75 to 100. Although this change relaxes the restrictions on which firms can operate under Subchapter S, it does not influence the relative tax advantages of the S corporation form and thus should not influence our results. 8 This restriction reduces the likelihood that alternative explanations related to non-tax features of C corporations are driving our results because S corporations cannot be foreign-owned and the limitation on the number of shareholders essentially prevents S corporations from being publicly traded. 9 We remove banks operating in Alaska, Washington DC, Delaware, Hawaii, Idaho, Massachusetts, Maine, New Hampshire, Rhode Island, Virginia, and Vermont. Including banks from these states does not influence our results. 10 The standard errors (0.0001 and 0.0002) for these mean tests of differences are relatively small because the interval is bounded between 0 percent and 100 percent. Thus, a 0.1 percent difference is statistically significant. Who Benefits from the Tax Advantages of Organizational Form Choice? 985 Q3 0.018 0.081 0.182 0.082 12.093 0.009 0.152 3.964 0.602 1.000 16.342 0.030 0.029 is the ratio of non-deposit ratio is the Lev 3.811 11.434

0.014 0.003 0.072 0.032 0.136 0.312 0.000 15.522 0.021 0.072 0.021 Median is the ratio of dividends to the absolute the to dividends of ratio the is Div Q1 is the lagged ratio of pre-tax income to total to income of pre-tax ratio lagged LagRoa is the 0.008 0.009 0.000 0.009 0.109 0.071 0.000 14.897 0.010 10.810 0.064 0.015 3.670 is the ratio of interest expense paid on deposits to year- of interest is the ratio DepRate 0.008 0.013 0.787 0.058 0.072 0.409 0.448 0.874 0.019 0.967 0.012 0.010 0.231 is the log of bank assets. bank of log Size is the Table 2 Table Standard Deviation Descriptive Statistics Descriptive is the percentage change in loans in the current year. year. current the in loans in change percentage the is LoanGrow 0.008 0.158 0.398 3.827 Mean 11.476 0.013 0.137 0.054 0.277 15.590 0.021 0.023 0.073 N 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 20,096 is the bank’s Tier 1 capital ratio. Tier Cap is the bank’s is the log of the ratio of total wages to the number of full-time equivalent employees. equivalent full-time of number the to wages total of ratio the of log the is Wage is the change in state-level gross domestic product. gross domestic state-level in change the GDP is is the ratio of interest income received on loans to year-end loan balance. on loans to year-end received income of interest is the ratio LoanRate is the ratio of non-performing loans to net loans. net to loans non-performing of ratio the is NPL LagRoa NPL LoanGrow Lev Cap SBank Pop GDP Size LoanRate

Div Variables Independent end deposits balance. balance. deposits end value of pre-tax income. SBank equals 1 for Subchapter S banks and 0 C banks. Pop is the log population in state which bank is headquartered. assets. liabilities to total assets. DepRate Wage Dependent Variables Dependent Notes: 986 National Tax Journal earnings as dividends (Div), while C banks pay only 28.2 percent (p<0.01). This result is consistent with S bank owners benefiting from the tax advantages of the S corporation form. Nevertheless, we conduct multivariate tests before making inferences.

B. Multivariate Tests Table 3 reports estimates of (1), where the dependent variable is indicated in each column heading for the relevant stakeholder outcome. In Column 1, the coefficient for SBank is insignificant whenLoanRate is the outcome. In Column 2, we find a negative and significant (p<0.01) coefficient forSBank when DepRate is the dependent variable, suggesting that S banks pay less to suppliers (i.e., depositors) than C banks. These results are not consistent with H1a or H2a, implying that customers and suppliers do not reap the tax benefits of S status. However, in Columns 3 and 4 we find positive and significant coefficients forSBank (p<0.01), suggesting S banks pay employees more in wages and owners more in dividends than C banks during the sample period. These results are consistent with H3a and H4a, suggesting that employees and owners, respectively, reap relatively more benefits of the S corporation form over the C corporation form. Table 4 reports estimates of (2), where we again indicate the outcome of interest in the column heading. In Column 1, the coefficients for SBank and SBank×Post are insignificant whenLoanRate is the outcome. In Column 2, we find negative and significant coefficients forSBank and SBank×Post when DepRate is the dependent variable, consistent with lower interest rates on deposits for S banks. However, these results are not consistent with the tax benefits of S corporations accruing to depositors because S corporations did not pay lower interest rates before JGTRRA. Instead, Columns 1 and 2 suggest that customers and suppliers, respectively, do not reap the benefits of S corporation status (i.e., the results are inconsistent with H1b and H2b). The results in Columns 3 and 4 indicate that while employees and owners of S banks reap benefits of the S corporation form (as predicted by H3b and H4b, respectively), these benefits declined after JGTRRA reduced the relative tax disadvantage of C corporations. First, the positive and significant coefficients for SBank in Columns 3 and 4 confirm that S banks pay employees more wages and owners more dividends, respectively, than C banks. Second, consistent with JGTRRA lowering the tax burden of both C and S banks, the coefficients forPost are positive and significant in Columns 3 and 4. These findings suggest that when statutory tax rates decline, sample banks increase both sala- ries and dividends. Third, consistent with S corporations realizing fewer tax benefits from JGTRRA than C corporations, the coefficients for SBank×Post are negative and significant. These results suggest that while S banks increased wages after JGTRRA, they did so to a lesser extent than C banks, because JGTRRA lowered their tax burden (i.e., personal ordinary tax rates) by a lesser amount than it lowered the dividend tax rate. Specifically, by summing the coefficients forPost and SBank×Post, we find that S banks did not increase their dividends after JGTRRA (p>0.76). Finally, consistent with Who Benefits from the Tax Advantages of Organizational Form Choice? 987

Table 3 Pooled Cross-Sectional Regression Results (Main Effects)

(1) (2) (3) (4) H1a H2a H3a H4a LoanRate DepRate Wage Div Coefficient Coefficient Coefficient Coefficient (RSE) (RSE) (RSE) (RSE) SBank −0.0003 −0.0006*** 0.0175*** 0.3828*** (0.0002) (0.0001) (0.0065) (0.0078) Pop −0.0003** −0.0009*** 0.0027 −0.0183*** (0.0001) (0.0001) (0.0034) (0.0040) GDP 0.0389*** −0.0319*** 0.4377*** −1.0275*** (0.0042) (0.0027) (0.1117) (0.1597) Size −0.0026*** 0.0000 0.0318*** −0.0258*** (0.0001) (0.0001) (0.0035) (0.0042) LagRoa 0.2753*** −0.0614*** −1.7016*** 8.9558*** (0.0159) (0.0086) (0.3404) (0.4099) NPL 0.1212*** 0.0312*** −0.5978*** 1.1709*** (0.0165) (0.0057) (0.1911) (0.2827) LoanGrow −0.0020*** −0.0007*** 0.0097*** −0.0059 (0.0005) (0.0002) (0.0026) (0.0055) Lev −0.0164*** 0.0108*** 0.2542*** −0.0868 (0.0016) (0.0012) (0.0517) (0.0578) Cap 0.0081*** −0.0053*** 0.0334 0.1965*** (0.0016) (0.0011) (0.0466) (0.0555) Constant 0.1147*** 0.0513*** 3.3381*** 0.7410*** (0.0021) (0.0014) (0.0637) (0.0721)

Year FE Included Included Included Included Adjusted R2 0.54 0.71 0.11 0.25 Observations 20,096 20,096 20,096 20,096 Notes: Robust standard errors (RSE) are clustered by bank (Petersen, 2009). Table 2 provides a list of variable definitions. Asterisks denote significance at the 1% (***), 5% (**), and 10% (*) levels. 988 National Tax Journal

Table 4 Difference-in-Differences Regression Results (Effects of JGTRRA) (1) (2) (3) (4) H1b H2b H3b H4b LoanRate DepRate Wage Div Coefficient Coefficient Coefficient Coefficient (RSE) (RSE) (RSE) (RSE) SBank −0.0003 −0.0005*** 0.0229*** 0.3937*** (0.0003) (0.0002) (0.0067) (0.0094) Post −0.0169*** −0.0170*** 0.1470*** 0.0183** (0.0002) (0.0001) (0.0034) (0.0077) SBank × Post 0.0000 −0.0003*** −0.0108*** −0.0217** (0.0002) (0.0001) (0.0035) (0.0109) Pop −0.0003** −0.0009*** 0.0027 −0.0183*** (0.0001) (0.0001) (0.0034) (0.0040) GDP 0.0389*** −0.0320*** 0.4350*** −1.0329*** (0.0042) (0.0027) (0.1118) (0.1596) Size −0.0026*** 0.0000 0.0318*** −0.0258*** (0.0001) (0.0001) (0.0035) (0.0042) LagRoa 0.2753*** −0.0618*** −1.7145*** 8.9300*** (0.0159) (0.0086) (0.3411) (0.4108) NPL 0.1212*** 0.0311*** −0.6008*** 1.1650*** (0.0165) (0.0057) (0.1911) (0.2821) LoanGrow −0.0020*** −0.0007*** 0.097*** −0.0059 (0.0005) (0.0002) (0.0026) (0.0055) Lev −0.0164*** 0.0108*** 0.2541*** −0.0870 (0.0016) (0.0012) (0.0517) (0.0578) Cap 0.0081*** −0.0053*** 0.0336 0.1968*** (0.0016) (0.0011) (0.0466) (0.0555) Constant 0.1147*** 0.0513*** 3.3370*** 0.7388*** (0.0021) (0.0014) (0.0637) (0.0721)

Year FE Included Included Included Included Adjusted R2 0.54 0.71 0.11 0.25 Observations 20,096 20,096 20,096 20,096 Notes: Robust standard errors (RSE) are clustered by bank (Petersen, 2009). Post equals 1 for observations in the post-JGTRRA period (after 2003); 0 otherwise. Table 2 provides a list of variable definitions. Asterisks denote significance at the 1% (***), 5% (**), and 10% (*) levels. Who Benefits from the Tax Advantages of Organizational Form Choice? 989

S banks enjoying a relative tax advantage over C banks, albeit to a lesser extent after JGTRRA, the sum of coefficients forSBank and SBank×Post are significantly positive when either Wage or Div is the dependent variable (p<0.05). To gauge the economic size of the employee wage results, we follow Kennedy (1981) and Donohoe and Knechel (2014). We multiply all coefficient estimates in Column 3 of Table 4 by the mean values for the corresponding variables, and incrementally add the coefficients for SBank, Post, and SBank×Post. We then compute the raw dollar value by exponentiating the summed value and multiplying by 1,000, as the figures in Call Reports are denominated in thousands of dollars. We find that annually, on average, S banks pay employees $43,363 whereas C banks pay $42,381, a difference of $981. After JGTRRA, S banks pay employees $49,690 and C banks pay $49,092, a difference of $597. Although these amounts seem modest, they are economically meaningful for at least two reasons. First, these values represent per employee annual averages. When we multiply these amounts by the average number of employees for banks in our sample (56.25), we find that the average S bank pays about $55,237 more in total wages than the average C bank, and $33,626 more after JGTRRA. Second, at the beginning of the sample period, the median wage in the United States was $21,767.11 The wage premiums related to the S corporation form thus represent a 4.5 percent average increase (2.7 percent after JGTRRA) in the median American’s compensation. We evaluate dividends in a similar fashion. We multiply the coefficients for SBank, Post, and SBank×Post in Column 4 of Table 4 by the average absolute value of pre- tax income, the scalar of Div ($2,437). We then multiply the result by 1,000. We find that the average S bank pays $270,618 more in dividends per year than the average C bank. After JGTRRA, all banks increase dividend payouts by approximately $12,578, on average; however, S banks increase dividends by $14,916 less than C banks.

V. ROBUSTNESS TESTS We evaluate the robustness of our results by re-estimating (2) after including firm fixed-effects. This adjustment mitigates correlated omitted variables bias because any confounding variable must be correlated with the sample bank over time (Wooldridge, 2010). Table 5 reports the results. Note that SBank is omitted from the model because it does not vary by firm and would otherwise be collinear with the firm-specific intercept. We find positive coefficients forPost when Wage or Div is the dependent variable (p<0.01), consistent with reductions in corporate tax rates benefiting employees and owners. In line with our main results, we find negative and significant coefficients for SBank×Post when Wage or Div is the dependent variable (p<0.01), consistent with JGTRRA lowering the tax burden on S corporations to a lesser extent than it did for C corporations. As before, we find insignificant coefficients whenLoanRate or DepRate is the dependent variable.

11 Social Security Administration, “Measures of Central Tendency for Wage Data,” www.ssa.gov/oact/cola/ central.html. 990 National Tax Journal

Table 5 Firm Fixed-Effects Regression

(1) (2) (3) (4) LoanRate DepRate Wage Div Coefficient Coefficient Coefficient Coefficient (RSE) (RSE) (RSE) (RSE) Post −0.0183*** −0.0189*** 0.1551*** 0.0449*** (0.0002) (0.0001) (0.0027) (0.0090) SBank × Post 0.0001 −0.0000 −0.0089*** −0.0319*** (0.0002) (0.0001) (0.0029) (0.0095) Pop 0.0057*** 0.0014 0.0270 −0.0579 (0.0018) (0.0009) (0.0291) (0.0960) GDP 0.0086*** 0.0013 0.0082 −0.1067 (0.0030) (0.0015) (0.0490) (0.1617) Size 0.0043*** 0.0043*** 0.0271*** −0.1748*** (0.0003) (0.0001) (0.0049) (0.0163) LagRoa 0.1336*** −0.0217*** −0.3858*** 6.8541*** (0.0088) (0.0043) (0.1452) (0.4797) NPL 0.0526*** 0.0173*** −0.2319*** 0.9050*** (0.0046) (0.0022) (0.0754) (0.2490) LoanGrow −0.0020*** −0.0006*** −0.0006 0.0133*** (0.0001) (0.0000) (0.0010) (0.0033) Lev −0.0228*** 0.0038*** 0.1384*** −0.0063 (0.0018) (0.0009) (0.0294) (0.0970) Cap 0.0326*** −0.0074*** 0.0486 −1.4926*** (0.0021) (0.0010) (0.0339) (0.1119) Constant −0.0580** −0.0338** 3.0072*** 3.4348** (0.0275) (0.0135) (0.4521) (1.4933)

Firm FE Included Included Included Included Year FE Included Included Included Included Adjusted R2 0.80 0.92 0.85 0.47 Observations 20,096 20,096 20,096 20,096 Notes: SBank is omitted as it does not vary by firm and would otherwise be collinear with the firm-specific intercept. Robust standard errors (RSE) are clustered by bank (Petersen, 2009). Post equals 1 for observations in the post-JGTRRA period (after 2003); 0 otherwise. Table 2 provides a list of variable definitions. Asterisks denote significance at the 1% (***), 5% (**), and 10% (*) levels. Who Benefits from the Tax Advantages of Organizational Form Choice? 991

Table 6 Single State Banks

(1) (2) (3) (4) LoanRate DepRate Wage Div Coefficient Coefficient Coefficient Coefficient (RSE) (RSE) (RSE) (RSE) SBank −0.0004 −0.0005*** 0.0218*** 0.3963*** (0.0003) (0.0002) (0.0068) (0.0095) Post −0.0169*** −0.0170*** 0.1465*** 0.0228*** (0.0002) (0.0001) (0.0034) (0.0078) SBank × Post 0.0000 −0.0003*** −0.0101*** −0.0254** (0.0002) (0.0001) (0.0036) (0.0111) Pop −0.0003** −0.0010*** 0.0020 −0.0172*** (0.0001) (0.0001) (0.0035) (0.0041) GDP 0.0393*** −0.0325*** 0.4091*** −1.0450*** (0.0043) (0.0028) (0.1127) (0.1641) Size −0.0027*** 0.0001 0.0328*** −0.0285*** (0.0001) (0.0001) (0.0036) (0.0045) LagRoa 0.2778*** −0.0616*** −1.6794*** 8.9904*** (0.0161) (0.0088) (0.3433) (0.4183) NPL 0.1179*** 0.0302*** −0.6571*** 1.1185*** (0.0163) (0.0057) (0.1866) (0.2824) LoanGrow −0.0020*** −0.0007*** 0.097*** −0.0053 (0.0005) (0.0002) (0.0026) (0.0054) Lev −0.0166*** 0.0112*** 0.2223*** −0.0813 (0.0016) (0.0012) (0.0523) (0.0595) Cap 0.0078*** −0.0053*** 0.0349 0.2062*** (0.0016) (0.0011) (0.0466) (0.0559) Constant 0.1154*** 0.0513*** 3.3395*** 0.7460*** (0.0021) (0.0014) (0.0642) (0.0739) Year FE Included Included Included Included Adjusted R2 0.54 0.71 0.11 0.25 Observations 19,476 19,476 19,476 19,476 Notes: The sample is restricted to banks operating only in a single state to ensure our main results are not due to differences in geographic scope among S and C banks. Robust standard errors (RSE) are clustered by bank (Petersen, 2009). Post equals 1 for observations in the post-JGTRRA period (after 2003); 0 otherwise. Table 2 provides a list of variable definitions. Asterisks denote significance at the 1% (***), 5% (**), and 10% (*) levels. 992 National Tax Journal

Table 7 No Large Banks

(1) (2) (3) (4) LoanRate DepRate Wage Div Coefficient Coefficient Coefficient Coefficient (RSE) (RSE) (RSE) (RSE) SBank −0.0003 −0.0005*** 0.0231*** 0.3938*** (0.0003) (0.0002) (0.0067) (0.0094) Post −0.0169*** −0.0170*** 0.1476*** 0.0181** (0.0002) (0.0001) (0.0034) (0.0077) SBank × Post 0.0000 −0.0003*** −0.0105*** −0.0218** (0.0002) (0.0001) (0.0035) (0.0109) Pop −0.0003** −0.0009*** 0.0029 −0.0179*** (0.0001) (0.0001) (0.0034) (0.0040) GDP 0.0394*** −0.0323*** 0.3958*** −1.0149*** (0.0042) (0.0027) (0.1099) (0.1599) Size −0.0026*** −0.0000 0.0305*** −0.0261*** (0.0001) (0.0000) (0.0035) (0.0043) LagRoa 0.2757*** −0.0622*** −1.7249*** 8.9570*** (0.0159) (0.0086) (0.3408) (0.4110) NPL 0.1211*** 0.0311*** −0.6009*** 1.1680*** (0.0165) (0.0057) (0.1912) (0.2822) LoanGrow −0.0020*** −0.0007*** 0.0097*** −0.0058 (0.0005) (0.0002) (0.0026) (0.0055) Lev −0.0164*** 0.0109*** 0.2483*** −0.0854 (0.0016) (0.0012) (0.0515) (0.0579) Cap 0.0082*** −0.0054*** 0.0290 0.1983*** (0.0016) (0.0011) (0.0467) (0.0556) Constant 0.1146*** 0.0514*** 3.3494*** 0.7350*** (0.0021) (0.0014) (0.0639) (0.0722)

Year FE Included Included Included Included Adjusted R2 0.54 0.71 0.11 0.25 Observations 20,064 20,064 20,064 20,064 Notes: The sample is restricted to banks that are no larger than the largest S bank in the sample (Beal Bank of Plano, TX, which holds approximately $5.5 billion in assets during 2002). This test ensures our main results are not driven by capital intensity. Robust standard errors (RSE) are clustered by bank (Petersen, 2009). Post equals 1 for observations in the post-JGTRRA period (after 2003); 0 otherwise. Table 2 provides a list of variable definitions. Asterisks denote significance at the 1% (***), 5% (**), and 10% (*) levels. Who Benefits from the Tax Advantages of Organizational Form Choice? 993

In addition, we examine whether differences in bank strategy or scope influence our results. S corporations face restrictions on issuing equity, which likely constrain their ability to expand nationally or pursue capital-intensive strategies. Thus, we limit our sample on two dimensions to ensure we are comparing relatively homogenous C and S banks. First, to consider whether our results for S corporation banks are due to differences in geographic scope, we estimate (2) on a subsample of banks that operate in a single state (19,476 observations).12 We report the results in Table 6. Similar to our main results, we find that S corporations pay employees more in wages and owners more in dividends (p<0.01 in both cases), on average. Further, after JGTRRA, all banks, on average, increase wages and dividends, but S corporations do so to a significantly lesser extent. Second, we remove any bank from the sample that is larger than the largest S bank in the sample (Beal Bank of Plano, TX, which holds about $5.5 billion in assets during 2002) to examine whether our results are driven by economies of scale. Table 7 reports the results. Using this subsample of 20,064 bank-years, we find evidence consistent with our main analyses. Finally, S banks are not geographically distributed on a random basis (Donohoe, Lisowsky, and Mayberry, 2015). Thus, S banks could face a different average level of competition than C banks, which might influence how banks shift the burden of taxation to different stakeholders. To examine this possibility, we estimate (1) and (2) after including two controls for competition: (1) the Herfindahl Index (HHI) of deposits at the state level; and (2) an indicator variable (Rural) equal to one for banks headquartered in a rural county (zero otherwise).13 We find evidence (untabulated) consistent with our main tests, suggesting that competition is not an alternative explanation. We also find that HHI has a positive (negative) and significant relation with Wage (Div), and that banks in rural counties (Rural) pay lower wages and more dividends than non-rural banks.

VI. CONCLUSION This study examines whether and the extent to which tax benefits from the S corporation organizational form accrue to customers, suppliers, employees, and/or owners. While similar along many non-tax dimensions, including limited personal liability, centralized management, and perpetual life, S corporations are subject to only one layer of tax (at the owner level) compared to C corporations that are subject to two layers of tax (at both the entity and owner levels). We find evidence consistent with employees and owners benefiting from the S corporation form through higher wages and dividends. However, customers and suppliers do not seem to benefit as deposit and loan interest rates do not systematically differ between the two organizational forms. Further, to

12 We obtain data on the locations of bank’s branches from the FDIC, “Summary of Deposits,” https://www5. fdic.gov/sod/. 13 We obtain data on state-level deposits from the FDIC Summary of Deposits. A bank is considered rural if it is not located in a Metropolitan Statistical Area (MSA), as defined by the U.S. Census Bureau. 994 National Tax Journal ensure that tax rather than unobservable non-tax attributes between the two corporate organizational forms explain our results, we validate our findings using the Jobs and Growth Tax Relief and Reconciliation Act of 2003 (JGTRRA), which dramatically reduced the tax advantage of S corporations over C corporations by lowering the dividend tax rate. Consistent with our primary inferences, we find that the wage premium and dividend payout rate for S corporations were attenuated following JGTRRA relative to that of C corporations. Our findings should be relevant to researchers, policymakers, and regulators. We find some evidence supporting a stated intention of the Small Business Job Protection Act of 1996 (SBJPA), which aimed to increased employee wages by allowing banks to change their organizational form from Subchapter C to Subchapter S status. However, we do not find support for the SBJPA’s goal of facilitating lower loan interest rates for bank customers. Our findings should also be informative to the current debate regarding corporate tax incidence. Importantly, our evidence suggests that a reduced (increased) corporate tax burden seems to benefit (harm) both employees and owners while leaving customers and suppliers unaffected.

ACKNOWLEDGMENTS and disclaimers We appreciate helpful comments from Paul Demeré, George Plesko, and John Robin- son (discussant). Special thanks to George Plesko for organizing the ATPI session held in Washington, DC as part of the NTA Spring Symposium on May 15, 2015, at which this paper was presented. Michael Donohoe and Petro Lisowsky gratefully acknowl- edge support from the PricewaterhouseCoopers Faculty Fellowship at the University of Illinois at Urbana-Champaign.

DISCLOSURES The authors have no financial arrangements that might give rise to a conflict of interest with respect to the research reported in this paper.

REFERENCES

Ablett, John, and Neil Hart, 2006. “Corporate Income Tax Reform: The Neglected Issue of Tax Incidence.” Economic Analysis and Policy 35 (12), 45–60. Allen, Eric J., and Sharat Raghavan, 2012. “The Impact of Non-Tax Costs on the Tax-Efficiency of Venture Capital Investments.” Manuscript. University of Southern California, Los Angeles, CA. Arulampalam, Wiji, Michael Devereux, and Giorgia Maffini, 2012. “The Direct Incidence of Corporate Taxes on Wages.” European Economic Review 56 (6), 1038–1054. Auerbach, Alan J., 2006. “Who Bears the Corporate Tax? A Review of What We Know.” In Po- terba, James M. (ed.), Tax Policy and the Economy, Volume 20, 1–41. MIT Press, Cambridge, MA. Who Benefits from the Tax Advantages of Organizational Form Choice? 995 aus dem Moore, Nils, and Tanja Kasten, 2009. “Do Wages Rise When Corporate Tax Rates Fall? Difference-in-Differences Analyses of the German Business Tax Reform 2000.” Ruhr Economic Papers Nr 532. Rheinisch-Westfälisches Institut für Wirtschaftsforschung (RWI), Essen, Germany. Ayers, Benjamin, C. Bryan Cloyd, and John Robinson, 1996. “Organizational Form and Taxes: An Empirical Analysis of Small Businesses.” Journal of the American Taxation Association 18 (Supplement), 49–67. Beatty, Anne, Philip G. Berger, and Joseph Magliolo, 1995. “Motives for Forming Research & Development Financing Organizations.” Journal of Accounting and Economics 19 (2), 411–442. Beaver, William H., Mary Lea McAnally, and Christopher H. Stinson, 1997. “The Information Content of Earnings and Prices: A Simultaneous Equations Approach.” Journal of Accounting and Economics 23 (1), 53–81. Berger, Allen N., and Christa H. S. Bouwman, 2013. “How Does Capital Affect Bank Performance During Financial Crises?” Journal of Financial Economics 109 (1), 146–176. Bertrand, Marianne, and Sendhil Mullainathan, 2003. “Enjoying the Quiet Life? Corporate Governance and Managerial Preferences.” Journal of Political Economy 111 (5), 1043–1075. Blankespoor, Elizabeth, Thomas Linsmeier, Kathy Petroni, and Catherine Shakespeare, 2013. “Fair Value Accounting for Financial Instruments: Does It Improve the Association between Bank Leverage and Credit Risk.” Accounting Review 88 (4), 1143–1177. Campbell, John L., James Chyz, Dan S. Dhaliwal, and William C. Schwartz Jr., 2013. “Did the 2003 Tax Act Increase Capital Investments by Corporations?” Journal of the American Taxation Association 35 (2), 33–63. Carroll, Robert, 2009. “Corporate Taxes and Wages: Evidence From the 50 States.” Tax Founda- tion Working Paper No. 8. Tax Foundation, Washington, DC. Clausing, Kimberly, 2013. “Who Pays the Corporate Tax in a Global Economy?” National Tax Journal 66 (1), 151–184. Donohoe, Michael P., 2015. “The Economic Effects of Financial Derivatives on Corporate Tax Avoidance.” Journal of Accounting and Economics 59 (1), 1–24. Donohoe, Michael P., and W. Robert Knechel, 2014. “Does Corporate Tax Aggressiveness Influ- ence Audit Pricing?” Contemporary Accounting Research 31 (1), 284–308. Donohoe, Michael P., and Gary A. McGill, 2011. “The Effects of Increased Book-Tax Difference Tax Return Disclosures on Firm Valuation and Behavior.” Journal of the American Taxation Association 33 (2), 35–65. Donohoe, Michael P., Petro Lisowsky, and Michael A. Mayberry, 2015. “Taxes, Competition, and Organizational Form.” Manuscript. University of Illinois at Urbana-Champaign, Champaign, IL. Erickson, Merle, and Shiing-wu Wang, 2000. “The Effect of Transaction Structure on Price: Evidence from Subsidiary Sales.” Journal of Accounting and Economics 30 (1), 59–97. 996 National Tax Journal

Erickson, Merle, and Shiing-wu Wang, 2007. “Tax Benefits as a Source of Merger Premiums in Acquisitions of Private Companies.” Accounting Review 82 (2), 359–387. Fuest, Clemens, Andreas Peichl, and Sebastian Siegloch, 2013. “Do Higher Corporate Tax Rates Reduce Wages? Micro Evidence from Germany.” IZA Discussion Paper No. 7390. The Institute for the Study of Labor (IZA), Bonn, Germany. Gentry, William M., 1994. “Taxes, Financial Decisions, and Organizational Form: Evidence from Publicly Traded Partnerships.” Journal of Public Economics 53 (2), 223–244. Goolsbee, Austan, 1998. “Taxes, Organizational Form, and the Deadweight Loss of the Corporate Income Tax.” Journal of Public Economics 61 (1), 143–152. Goolsbee, Austan, and Edward Maydew, 2002. “Taxes and Organizational Form: The Case of REIT Spin-offs.” National Tax Journal 55 (3), 441–456. Gravelle, Jennifer, 2013. “Corporate Tax Incidence: Review of General Equilibrium Estimates and Analysis.” National Tax Journal 66 (1), 185–214. Guenther, David A., 1992. “Taxes and Organizational Form: A Comparison of Corporations and Master Limited Partnerships.” Accounting Review 67 (1), 17–45. Hanlon, Michelle, and Shane Heitzman, 2010. “A Review of Tax Research.” Journal of Account- ing and Economics 50 (2–3), 127–178. Harberger, Arnold C., 1962. “The Incidence of the Corporation Income Tax.” Journal of Political Economy 70 (3), 215–240. Hein, Scott E., Timothy W. Koch, and S. Scott MacDonald, 2005. “On the Uniqueness of Com- munity Banks.” Economic Review 90 (1), 15–36.

Internal Revenue Service (IRS), 2013a. Instructions for Form 1120S (2013). Internal Revenue Service, Bristol, VA. Internal Revenue Service (IRS), 2013b. “SOI Tax Stats – Numbers of Returns Filed by Type of Return – IRS Data Book Table 2.” Internal Revenue Service, Bristol, VA, http://www.irs.gov/ uac/SOI-Tax-Stats-Numbers-of-Returns-Filed-by-Type-of-Return-IRS-Data-Book-Table-2. Jin, Justin Y., Kiridaran Kanagaretnam, Gerald J. Lobo, and Robert Mathieu, 2013. “Impact of FDICIA Internal Controls on Bank Risk Taking.” Journal of Banking and Finance 37 (2), 614–624. Kennedy, Peter E., 1981. “Estimation with Correctly Interpreted Dummy Variables in Semiloga- rithmic Equations.” American Economic Review 71 (4), 801. Liu, Li, and Rosanne Altshuler, 2013. “Measuring the Burden of the Corporate Tax Under Im- perfect Competition.” National Tax Journal 66 (1), 215–238. Luna, LeAnn, and Matthew N. Murray, 2010. “The Effects of State Tax Structure on Business Organizational Form.” National Tax Journal 63 (4), 995–1021. Who Benefits from the Tax Advantages of Organizational Form Choice? 997

MacKie-Mason, Jeffrey K., and Robert H. Gordon, 1997. “How Much Do Taxes Discourage Incorporation?” Journal of Finance 52 (2), 477–505. Petersen, Mitchell A., 2009. “Estimating Standard Errors in Finance Panel Data Sets: Comparing Approaches.” Review of Financial Studies 22 (1), 435–480. Petroni, Kathy R., and Douglas A. Shackelford, 1995. “Taxation, Regulation, and the Organiza- tional Structure of Property-Casualty Insurers.” Journal of Accounting and Economics 20 (3), 229–253. Roberts, Michael R., and Toni M. Whited, 2013. “Endogeneity in Empirical Corporate Finance.” In Constantinides, George M., Milton Harris, and Rene M. Stulz (eds.), Handbook of the Econom- ics of Finance, Volume 2, 493–572. Elsevier, Amsterdam, Netherlands. Ryan, Stephen G., 2007. Financial Instruments and Institutions: Accounting and Disclosure Rules, 2nd Edition. John Wiley & Sons, Hoboken, NJ. Sahni, Balbir S., and T. Mathew, 1976. The Shifting and Incidence of the Corporation Income Tax. Rotterdam, The Netherlands: Rotterdam University Press. Scholes, Myron S., Mark A. Wolfson, Merle M. Erickson, Michelle L. Hanlon, Edward L. Maydew, and Terrence J. Shevlin, 2014. Taxes and Business Strategy, 5th Edition. Prentice Hall, Upper Saddle River, NJ. Sebold, Frederick D., 1979. “The Short-Run Shifting of the Corporation Income Tax: A Simul- taneous Equation Approach.” Review of Economics and Statistics 61 (3), 401–409. Shelley, Marjorie K., Thomas C. Omer, and T. J. Atwood, 1998. “Capital Restructuring and Ac- counting Compliance Costs: The Case of Publicly Traded Partnerships.” Journal of Accounting Research 36 (2), 365–378.

Shevlin, Terry, 1987. “Taxes and Off-Balance Sheet Financing: Research and Development Limited Partnerships.” Accounting Review 62 (3), 480–509. Terando, William D., and Thomas C. Omer, 1993. “Corporate Characteristics Associated with Master Limited Partnership Formation.” Journal of the American Taxation Association 15 (1), 23–46. Wooldridge, Jeffrey M., 2010. Econometric Analysis of Cross Section and Panel Data. MIT Press, Cambridge, MA. Zhang, Yi, Kathleen A. Farrell, and Todd A. Brown, 2008. “Ex-Dividend Day Price and Volume: The Case of the 2003 Dividend Tax Cut.” National Tax Journal 61 (1), 105–127.