CAPITAL INVESTMENT EFFECTS OF DIVIDEND IMPUTATION, …
CAPITAL INVESTMENT EFFECTS OF DIVIDEND IMPUTATION, CAPITAL GAINS TAX AND THE INVESTMENT TAX CREDIT
by
Ervin L. Black*
Assistant Professor
Joseph Legoria**
Assistant Professor
Keith F. Sellers*
Associate Professor
*Department of Accounting
College of Business
University of Arkansas
Fayetteville, AR 72701
(501) 575-6803
email: eblack@comp.uark.edu
email: sellers@comp.uark.edu
**School of Accountancy
College of Business and Industry
Mississippi State University
Mississippi State, MS 39762-5661
(601) 325-1634
email: jlegoria@cobilan.msstate.edu
This paper has benefited from comments of Terry Shevlin, David Burgstahler and other workshop participants at the University of Washington and participants at the Tenth Asian-Pacific Conference on International Accounting Issues.
CAPITAL INVESTMENT EFFECTS OF DIVIDEND IMPUTATION, CAPITAL GAINS TAX AND THE INVESTMENT TAX CREDIT
Abstract
We examine the effects of specific tax reforms on corporate capital investment in New Zealand, Australia, and Canada. The empirical findings indicate that: (1) tax reform in each of the three countries stimulated corporate capital investment, (2) tax reform altered how dividend payout ratios impacted capital investment in Canada but not in the other two countries, and (3) tax reform altered how capital intensity influenced investment in New Zealand and Australia but not Canada. Additional analyses are performed by dividing the samples into portfolios based on historical dividend payout policies and capital intensity. The findings indicate that tax reform impacted corporate investment differently depending on a firm’s dividend payout and capital intensity. In summary, we demonstrate the impact of specific tax reforms in three countries on corporate investment. Our findings suggest that policy makers can make more informed decisions regarding tax policy as it affects capital investment by examining the impact of tax reforms in other countries.
CAPITAL INVESTMENT EFFECTS OF DIVIDEND IMPUTATION, CAPITAL GAINS TAX AND THE INVESTMENT TAX CREDIT
INTRODUCTION
Economic theory has long held that the level of business investment in fixed assets is a critical determinant of business output and empirical research has clearly demonstrated this relation.[1] Based on this association, the stimulation of capital investment has been one of the principal objectives of tax policy in the U.S. for decades. Historically, Congress has attempted to stimulate capital investment through direct incentives such as the investment tax credit (ITC) or generous accelerated depreciation allowances. Surprisingly, research has shown that these incentives produce relatively small or no incremental increase in capital investment.
Macroeconomic theory asserts that tax policy should encourage investment in capital assets if the tax system lowers the marginal cost of investment. This theory has been the foundation of most investment incentives such as the ITC. For example, by directly lowering the cost of a fixed asset, the ITC should result in more investment in fixed assets. One weakness with such a targeted incentive is that all qualifying investment generates a credit, not just incremental investments. Furthermore, it is clear that at least some targeted tax incentives have led to sub-optimal investment decisions. An alternative method of lowering the cost of capital and thus stimulating investment for businesses is to simply lower the overall tax burden on business profits.
The purpose of this study is to determine the effects of specific income tax reforms on corporate fixed investment. Specifically, we examine the incremental effects of dividend imputation, capital gains taxes, and investment tax credits on investment. Evidence that these specific tax reforms result in increased or decreased investment is relevant for future tax policy debates in the U.S. and elsewhere. Ongoing debates over the merits of alternative tax reforms, such as capital gains tax preferences and dividend imputation, indicate that this is a timely and relevant research topic.
In order to determine the relative and incremental effects of various tax rates and incentives, tax policy changes in three countries are examined. In 1987, New Zealand adopted dividend imputation and significantly lowered corporate tax rates. Dividend imputation has the effect of eliminating the double tax on corporate profits. The second country examined in the study is Australia. In 1987, Australia adopted a tax reform plan similar to New Zealand’s. However, Australia also instituted for the first time a capital gains tax on sales of corporate stocks. This has the effect of lowering the combined tax on corporate profits which were distributed, but raising the combined rates on undistributed profit.
The third country to be examined is Canada. In 1972, Canada adopted a tax reform package similar to the one described above for Australia. Specifically, Canada addressed the double tax on dividends through a dividend-imputation plan while instituting a first time tax on capital gains. However, Canada also simultaneously adopted an ITC. It is important to note that all three countries were attempting to stimulate investment through their respective tax reforms.[2] While all three countries adopted similar forms of dividend imputation in an attempt to eliminate the double taxation of distributed corporate profits they differed in their
approaches on taxing capital gains and for providing direct investment incentives.
This combination of similar and different tax changes among the three countries offers an ideal setting for our study. By treating the tax changes of these countries as experiments in tax and investment policy, we can compare the relative and incremental effects of capital gain taxes and investment tax credits in conjunction with the adoption of dividend imputation. Economic conditions differed over time and among countries and general economic conditions certainly impact investment in fixed assets. However, we control for various non-tax variables that affect corporate investment in property, plant and equipment.
Our empirical findings indicate that: (1) tax reform in each of the three countries stimulated corporate investment, (2) tax reform altered how dividend payout ratios impacted investment in Canada but not in the other two countries, and (3) tax reform altered how capital intensity influenced investment in New Zealand and Australia but not Canada. In summary, we find evidence that specific tax reforms introduced in each of the three countries impacted corporate investment. Thus, our findings suggest that policy makers can make more informed decisions regarding tax policy as it affects capital investment by examining the impact of tax reforms in other countries.
In the following sections we examine prior research, develop hypotheses, discuss research design, provide results, and present conclusions and implications from our findings.
PRIOR RESEARCH ON TAXES AND INVESTMENT
Prior research on investment and taxes has generally relied on either the "user cost of capital" approach developed by Jorgensen (1963) and Hall and Jorgensen (1967) or Tobin's (1969) q-theory. However, researchers have had only limited success in linking tax changes to changes in investment using these models.[3] In fact, simple time series models or ad hoc models using variables such as output, cash flow, profits, and sales tend to predict total investment better than tax changes or their anticipated effects to cost of capital.[4]
Recently, studies utilizing firm specific panel data and/or specific tax reforms have been more successful in demonstrating a shift in corporate investment in response to tax changes. For example, Rosacher et al. (1993) examined how effective the ITC was over the period from its original enactment in 1962 to its repeal in 1986. The authors divide their sample into a test group consisting of firms qualifying for the ITC and a control group consisting of firms that do not qualify for the ITC. Using a univariate Box-Jenkins interrupted analysis, they find that ITC enactments and rate enhancements have a positive effect on investment whereas repeals have a negative effect on investment. Ayres (1987) employed financial capital markets research methodology in testing the effects of the ITC on security returns. She finds a significant association between abnormal security prices and the amount of ITC received (lost) do to changes in the ITC. Her results indicate that changes in the ITC result in a reallocation of capital among firms.
Moore et al. (1987) test whether tax advantages offered by various states in the U.S. are important in the decision of foreign corporations to invest in one state rather than another. Their findings indicate that tax structures relying on the unitary method of accounting
significantly impact the amounts of investment while corporate tax rates do not.[5] However, in a subsequent article Swenson (1989) used an experimental economics approach to test whether taxation impacted investment. He found that progressive tax regimes tend to decrease demand for fixed assets while tax credits resulted in increased demand for those assets. In addition, Vines et al. (1994) found that states with strong business interests (e.g. large number of firms and high average business income per firm) tend to have lower state corporate tax rates. Cassou (1997) studied the impact of tax policy on the flow of foreign investment between the U.S and other countries. His findings indicate there is a significant negative relation between U.S corporate tax rates and the level of foreign direct investment. These findings are consistent with Moore et al. (1987) and provide further evidence that tax policy is an important consideration for foreign firms when deciding where to invest.
Kern (1994) tested the impact of the Economic Recovery Act of 1981 (ERTA) on corporate investment. Under the ERTA, long-lived assets received greater benefits than short-lived assets. Kern divided her sample into three portfolios based on the after-tax benefits received from ERTA. The findings of her study indicate that firms receiving the greatest tax benefits exhibited the largest change in investment patterns and suggest that tax policy can have an effect on investment. However, Courtenay et al. (1989) found evidence that the ERTA disturbed the degree of neutrality in the tax law existing between capital intensive and non-capital intensive firms. More specifically, they found that only capital intensive firms
demonstrated significantly positive abnormal returns during their test period surrounding the passage of ERTA. They concluded that all ERTA may have accomplished was a reallocation of resources from non-capital intensive firms to capital intensive firms. In addition, Swenson (1987) determined that the Accelerated Cost Recovery System (ACRS) passed under ERTA was non-neutral during periods of high inflation.[6]
In 1984, the United Kingdom passed tax reform legislation which reduced corporate tax rates and lengthened the asset lives for many depreciable assets. Moon and Hodges (1989) examined how the 1984 U.K. tax reform affected the after-tax cost of capital facing firms. Their analysis indicates that the 1984 U.K. tax reform resulted in many investments in plant and equipment becoming less attractive than they were before. Morgan (1992) surveyed the largest U.K. firms to determine how sensitive these firms were to the 1984 U.K. tax reform. He found that most of the firms sensitive to the 1984 tax changes (e.g. firms with high marginal tax rates) would have scaled back their investment rather than have increased it.
In an examination of the Tax Reform Act (TRA) of 1986, Cummins and Hassett (1992) find a significant relation between the cost of capital and the level of investment in equipment and structures. Auerbach et al. (1991) find that the TRA of 1986 resulted in less investment than what was predicted based on investment behavior from 1953 to 1985. Cummins et al. (1994) found that after every tax reform enacted in the United States since 1962, the level of investment changed. In a subsequent article, Cummins et al. (1996) examine various tax
reforms in 14 countries and find evidence that taxes, in general, can be linked to changes in investment.
Finally, Kinney and Trezevant (1993) analyzed whether taxes impact the timing of capital expenditures. More specifically, they argue that the present value of investment-related tax shields is greater if a depreciable asset is purchased and placed in service in the current year compared to the subsequent year. As a result, they predict greater capital expenditures are made in the fourth quarter of the current year, as opposed to the first of quarter of the following year. They find support for this prediction as their results indicate firms make greater capital expenditures in the fourth quarter of current year rather than first quarter of the next year.
In summary, prior research has determined that the long suspected link between taxes and investment does exist. Unfortunately, no previous research has successfully demonstrated the investment effects of specific tax reforms such as reduction of corporate tax rates or elimination of the double tax on dividends. Thus, while tax policy-makers know that their decisions might impact the level of investment, they do not have information as to which specific tax changes result in changes in actual investment behavior.
The objective of our research study is to increase understanding of the impact of specific tax changes; i.e., changes in the structure of the tax system on distributed and non-distributed corporate earnings and their effect on corporate investment behavior. By examining the investment effects in countries that have implemented specific tax reforms, we derive implications of the potential effects of similar changes in the U.S. and other countries.[7]
HYPOTHESIS DEVELOPMENT
Taxes, the Cost of Capital, and Capital Investment
The purpose of this section is to demonstrate the various linkages between components of a capital investment model and the various tax changes examined in this study. Firms should accept additional investment opportunities if the net present value of the marginal investment is positive. Assuming the firm has sufficient capital resources, it would invest if:
(1)
where: CFt = the net cash flows during period t, and
k = the investor’s required rate of return
Therefore, taxes and tax changes impact the investment decision to the extent that they impact the timing of projected net cash flows (t), the amount of projected net cash flows (CF), discount rates (k), or a combination of these factors. Corporate level taxes clearly affect the timing and amount of cash flows. Shareholder level taxes affect corporate investment decisions less directly, through their impact on the corporate cost of capital. In the absence of shareholder level taxes, the corporate cost of equity is equal to the expected rate of return to shareholders in the form of dividends and appreciation. When shareholder level taxes are imposed, the corporate cost of equity capital and total shareholder return differ by the amount of total shareholder level taxes. Assuming that a corporation’s risk adjusted required rate of return on marginal investments is equal to the corporate cost of capital, the relationship between individual shareholder taxes and the cost of capital under a classical double tax system can be represented as:
(2)
where:
Ce = corporate cost of equity,
Ts = total shareholder level taxes,
Rs = total expected after-tax return to shareholders,
Rd = expected shareholder return in the form of dividends,
Td = marginal tax rates on dividend income, and
Rcg = expected shareholder return in the form of share appreciation (capital gain)
The corporation’s cost of equity capital differs from the shareholders’ after-tax return by the amount of the shareholder-level tax, in this case equal to Rd*Td.[8] Prior to the tax reforms examined in this study, neither Australia, Canada nor New Zealand imposed a tax on realized capital gains. Thus, Rcg is equal to a shareholder’s pre and post-tax return in the form of share appreciation. It should be noted that, in most instances where statutory rates for Td are non-zero, its marginal value is non-observable. Various tax clienteles will naturally concentrate their investments in appropriate stocks based on their tax position and the expected amount and form of the return. This clientele effect, combined with variations in the timing of dividend distributions, make the present value of the tax impossible to estimate.
New Zealand
Dividend imputation modifies the tax on dividends by attaching a credit, equal to the taxes already paid by the corporation on behalf of the dividend, to dividend distributions. As outlined in the appendix, if individual tax rates exceed corporate tax rates, the imputation credit will only partially offset the individual level taxes. While some view this as only partially reducing the classic double tax, the resulting cumulative tax burden on distributed corporate earnings is equal to only the higher individual tax rate. In the case where the corporate tax rate exceeds the tax rate of the recipient shareholder and the imputation credit can be fully utilized, the resulting cumulative tax burden is once again equal to the individual shareholder’s marginal tax rate. However, the after-tax dividend return, as viewed by the shareholder, is actually higher than the pre-tax return. These conclusions are evident in the following when we add the tax benefit of the imputation credit:
Rs = Rd – (Rdg * Td) + (Rdg * Tc) + Rcg (3)
where:
Rdg = the dividend return paid to shareholders, “grossed-up” by the amount of
corporate level tax paid on account of the dividend, and
Tc = the corporate tax rate paid on the grossed-up distributed earnings.
Equation (3) captures the cost of capital in New Zealand after the tax reform and accurately reflects the effects of the imputation credit when the shareholder has a positive tax rate and sufficient income. If the recipient has insufficient income, some or all of the imputation credit will be lost. Subtracting equation (2), which measures the after-tax return to shareholders prior to dividend imputation from equation (3), the after-tax return after the tax reform, reveals that shareholders’ after-tax return in New Zealand changed by an amount equal to (Rdg * Tc) - (Rdg * Tc * Td). When shareholder tax rates (Td) fall between zero and 100 percent, this will result in a positive change in shareholder’s after-tax returns. Thus, the tax reform in New Zealand clearly enhanced shareholder after-tax returns for non-tax exempt shareholders.
Tax clientele theory, however, suggests that the higher rate of after-tax return arising from lower explicit taxes will be offset by higher implicit taxes. Investors will bid the price of dividend-paying stocks up until the after-tax return returns to its equilibrium state. This market adjustment results in an unambiguous reduction in the corporate cost of additional equity financing, and a corresponding reduction in required rates of return on corporate capital investments.[9] This leads to our first hypothesis:
H1: The introduction of a dividend imputation tax system in New Zealand significantly
increased corporate capital investment.
Australia
In Australia, where the tax reform also included the imposition of a new tax on capital gains, the effect on shareholder returns of the new tax is shown by modifying equation (3) as follows:
Rs = Rd – (Rdg*Td) + (Rdg*Tc) + Rcg*(1-Tcg) (4)
where: Tcg = the tax rate on capital gains.
As with Td, the value for the capital gain tax rate (Tcg) is unobservable. Like the U.S., both Canada and Australia defer the tax on share appreciation until gains are realized by the shareholder. The period of this deferral is determined by the shareholder, and can even extend beyond the life of an individual shareholder. Once again, by subtracting equation (2) from equation (4), one can determine that the after-tax return for shareholders in Australia increased by an amount equal to (Rdg*Tc - Rdg*Td*Tc) - (Rcg*Tcg). Conclusions as to the net effect of this tax reform on shareholder’s after-tax returns, and thus the cost of corporate equity capital, are ambiguous. The change arising from dividend imputation (Rdg * Tc – Rdg * Td * Tc) increased shareholder returns while the new capital gains tax reduced shareholder returns by an amount equal to Rcg * Tcg. As we cannot observe values for the marginal shareholder-level tax rates Td and Tcg, our hypothesis for Australia is non-directional. To determine the net impact of dividend imputation and the imposition of the capital gains tax on shareholder after-tax returns, and thus corporate cost of equity capital and corporate capital investment, we empirically test the following:
H2: The introduction of a dividend imputation tax system and a capital gains tax
significantly impacted corporate capital investment in Australia.
Because of the counteracting effects of dividend imputation and capital gains tax we cannot predict the directional impact on capital investment. Therefore, we formally test the following hypotheses.
H2(a): if the positive effects of dividend imputation on shareholder returns exceeded the
negative effects of capital gains taxes on shareholder returns, corporate capital investment in Australia increased after tax reform.
H2(b): if the negative effects of capital gains taxes on shareholder returns exceeded the
positive effects of dividend imputation on shareholder returns, corporate capital investment in Australia decreased after tax reform.
H2(c): if the negative effects of capital gains taxes on shareholder returns equally offset the
positive effects of dividend imputation on shareholder returns, corporate capital investment in Australia was unaffected after tax reform.
Canada
The Canadian tax reform examined in this study was similar to that of Australia’s in that it implemented both dividend imputation and a new capital gains tax. For this study, a critical difference between the Canadian tax reform and that of Australia is that the Canadian tax act also provided a new investment tax credit. Unlike changes in shareholder level taxes, investment tax credits directly influence corporate capital investment by enhancing net cash flows from investments (CF in equation 1), rather than the cost of capital. Thus, as with Australia, the Canadian tax reform introduced provisions with both positive and negative implications for corporate capital investment. We examine the net effects of these various provisions on capital investment by testing the following hypothesis:
H3: The introduction of a dividend imputation tax system, a capital gains tax and an
investment credit significantly impacted corporate capital investment in Canada.
As in Australia, because of the counteracting effects of dividend imputation, the capital gains tax, and an investment tax credit, we cannot predict the direction of the impact on investment in Canada. Thus, we formally test the following hypotheses.
H3(a): if the positive effects of dividend imputation and the investment tax credit exceeded
the negative effects of capital gains taxes, corporate capital investment in Canada increased after tax reform.
H3(b): if the negative effects of capital gains taxes exceeded the positive effects of dividend
imputation and the investment credit, corporate capital investment in Canada decreased after tax reform.
H3(c): if the negative effects of capital gains taxes equally offset the positive effects of
dividend imputation and the investment credit, corporate capital investment in Canada is unaffected by tax reform.
Because we predict offsetting effects of the tax changes, examination of an aggregate sample of Australian and Canadian corporations can only reveal the net impact, if any, of the tax reforms of those countries. By segregating corporations by the form of return provided to shareholders, it may be possible to draw inferences as to the effects of the specific tax changes, individually and interactively. For example, assume Australian shareholders invested in a corporation with the expectation of receiving returns solely in the form of dividends. In this case, the change in shareholder returns, and thus the corporation’s cost of equity capital, would only be impacted by dividend imputation. Expected after-tax returns in the form of share appreciation remained zero and played no part in the market’s valuation of those shares. For a portfolio of such firms, the effects of the capital gains tax are assumed to be zero and any observed change in corporate investment could be attributable to dividend imputation. Similarly, for Australian firms with expected returns solely in the form of share appreciation, only the imposition of the new capital gains tax should impact the cost of equity.
In an effort to disaggregate the effects of dividend imputation from capital gains taxes and investment tax credits, we divide the sample from each country into four portfolios based on dividend payout policies and capital intensity.[10] As shown in Table 1, the four portfolios, I, II, III, and IV, include low dividend payout/low capital intensity, high dividend payout/low capital intensity, low dividend payout/high capital intensity, and high dividend payout/high capital intensity firms, respectively. These portfolios are partitioned based on the median values of dividend payout and capital intensity ratios. Table 1 describes expectations of how tax reform impacts each of the four portfolios in each of the three countries.
RESEARCH DESIGN
Variable Selection and Measurement
Dependent Variable
Investment (INV): An ideal measure of capital investment would be the amount of cash spent on capital expenditures obtained from the cash flow or funds statement. While Canada requires firms to disclose this information, this measure is not available to us for firms from New Zealand and Australia. Thus, to measure investment consistently across the three countries, we follow Kinney and Trezevant (1993) and measure capital investment as the change in gross property, plant, and equipment (GPPE). In addition, we add a firm’s annual R&D expenditures to its change in GPPE for each year. Most countries require firms to
expense R&D in the current period because it is difficult to determine the future revenues that will result from current R&D costs. However, Lev and Sougiannis (1996) find that the market capitalizes R&D costs and this capitalization provides useful information about future earnings. Therefore, we feel that a measure of investment that captures both a firm’s investment in fixed assets and R&D more accurately reflects a company’s investment decisions.
We scale investment by total sales consistent with Kern (1994) to obtain a measure of investment to control for inflation and growth. Therefore, our measure of investment is the change in GPPE plus R&D divided by sales.[11]
Independent Variables
Independent variables for this study can be categorized as either tax variables or firm specific variables. The tax variables represent the variables of interest for this study, and are designed to capture the effects of the various tax changes. Firm specific variables are used to control for other determinants of corporate capital investment behavior, primarily the firm’s investment opportunities.
Tax Variables
A common problem in the analysis of tax law changes is that legislatures tend to make many, often significant, tax changes simultaneously. Disentangling the effects of the various changes presents a challenge
Tax Act (TREF): The variable of primary interest in this study indicates the effect of tax reform in the various countries. Thus, TREF is set to zero in years preceding the effective date of tax reform and one in years in which dividend imputation is in effect. In New Zealand, this variable measures only the impact of dividend imputation. In Australia, this variable also measures any impact of implementing capital gains taxes. Finally, in the Canadian sample this variable will capture the effects of dividend imputation, the imposition of a capital gains tax, and the implementation of an investment tax credit.
Statutory Tax Rate (RATE): During the years surrounding the adoption of reform, each country significantly reduced corporate and individual income tax rates. Because both tax rates theoretically impact corporations’ investment decisions, ideally, both individual and corporate rates would be included in the model to account for these effects. However, individual and corporate income tax rates exhibit a high empirical correlation over the sample period of this study, and the presence of both variables would introduce multicollinearity into the model. Thus, for this study RATE is defined as the maximum of either the federal corporate tax rate or the individual tax rate, expressed as a percentage. Since RATE proxies for both individual and corporate tax rates, the sign of its effect on capital investment is ambiguous.[12]
Firm Specific Variables
Net Operating Loss (DNOL): As a firm’s taxable income decreases, the possibility of incurring a tax net operating loss (NOL) increases, resulting in a lower projected marginal tax rate. Firms with relatively low taxable income may thus respond differently to tax reforms such as dividend imputation and lower tax rates. For example, imputation credits may be
unavailable if a company is not paying sufficient corporate level taxes, greatly diminishing any impact of dividend imputation. Similarly, the incentives provided by traditional tax shields or lower statutory rates decrease as a firm’s marginal tax rate decreases. In order to capture the effects of low marginal tax rates, we use a dummy variable set equal to one if the firm is likely to have a net operating loss, zero otherwise. A firm is deemed likely to have a net operating loss if it has had one in the immediate past. We determine this by examining each of the last five years. DNOL is equal to one if the firm had a net operating loss in any one of the five years, zero otherwise.[13]
Book-to-Market Equity Ratio (BKMKT): Perhaps the most widely used estimate of a firm’s investment opportunity set is the ratio of the market value of equity to the book value of equity (and its inverse). Collins and Kothari (1989) assert that the market-to-book ratio captures the difference between a firm’s return on both existing and future assets and its required rate of return on equity. Thus, a lower ratio indicates greater investment opportunities which are expected to yield returns in excess of the required rate of return.[14] Since this is a control variable, we also try other theoretically sound proxies of a firm’s investment opportunities including the earnings to price ratio and measures of variability (as used by Christie (1989) and Smith and Watts (1992).[15]
Debt to Equity Ratio (D/E): In an efficient market, is reasonable to assume that capital is provided to firms with better investment opportunities. However, as discussed above, each of our direct proxy measures of investment opportunity are sensitive to the capital structure of the firm.
Cash Flow Ratio (CF): In addition to externally generated capital a firm generates capital internally. The cash flow ratio, measured as operating cash flow divided by net sales, is a measure of the firm’s ability to internally finance its capital expenditures/investments. A firm with limited ability to internally fund investment opportunities is less likely to be able to increase investment. Conversely, a firm with a large cash flow ratio would be able to increase investment more easily.
Dividend Payout Ratio (DP): Shareholders receive returns on corporate investments in the form of share appreciation and dividends. When these two forms of return are taxed differentially, as they are in each of the tax systems examined in our study, shareholders are attracted to appropriate clienteles as a means of maximizing their total after-tax return. Dividend imputation is designed to eliminate the effects of the traditional double-tax on corporate earnings. However, many firms never pay cash dividends and are thus not directly affected by the double-tax system. It is anticipated that these firms will experience little or no impact from dividend imputation while traditionally high-dividend paying firms will be more strongly impacted. On the other hand, low dividend paying firms are more strongly impacted by capital gains taxes. Thus, we expect some interaction between a firm’s dividend payout and its response to a package of tax reform initiatives which include dividend imputation and a capital gains tax. We test for the difference in the coefficient of this variable before and after the tax reform acts.
Capital Intensity (CINT): The decision to invest in fixed assets (Property, Plant, and Equipment) is dependent not only on the tax system, but also on the firm’s industry or type of business. Firms in service industries that are not capital intensive would not be expected to significantly adjust their investment based on tax effects. Therefore, it is expected that capital investment by firms with high capital intensity are more affected by tax reforms than are firms with relatively lower capital intensity. Capital intensity is measured as the ratio of fixed assets to total assets. We test for the difference in the coefficient of this variable before and after the tax reform acts.
Size: Size of the firm also affects its ability to finance additional investment and is used as a control variable. Size is measured as the natural log of total assets.
Lagged Investment (LINV): Over time, the level of investment should tend to correlate with certain industry characteristics. Also, firms in various life stages may be experiencing relatively higher or lower levels of capital investment. In both cases, capital investment for a firm in any year should be correlated with investment in the succeeding year. Thus, we included a measure of capital investment for the prior year, scaled by total sales to be consistent with our dependent variable. Because it is possible that capital investment might display a cyclical nature, with large investments followed by periods of lower investment, we do not predict a sign for this coefficient.
Data Selection
We examine pooled cross-section firm-year data from three countries that experienced major tax reform believed to affect corporate capital investment: New Zealand, Australia, and Canada. In 1987, New Zealand adopted dividend imputation and significantly lowered corporate tax rates. Also in 1987, Australia adopted a dividend imputation plan similar to that of New Zealand. However, Australia also instituted for the first time a capital gains tax on sales of corporate stocks.[16] The third country to be examined is Canada. In 1972, Canada adopted a tax reform package similar to the one described above for Australia. However, Canada also simultaneously adopted an ITC. The tax reform acts in each of these countries is summarized in the Appendix.
Financial statement information for New Zealand and Australia is obtained from the Compustat Global Vantage database for 1982 through 1991. The Canadian data is obtained from Canadian Compustat for 1968-1977. Firm-year observations are deleted if the necessary data to calculate regression variables are not available. These data allow us to observe several years before and after tax reform in each country. After deleting observations due to lack of data, the New Zealand sample consists of 158 firm-year observations, Australia has 1,010 firm-year observations, and the Canadian sample has 1,776 observations.
Descriptive Statistics
Maximum corporate and individual tax rates in effect during the periods examined in the study are shown in Table 2. Although there were annual variations, rates for both individuals and corporations were generally lower in years after the tax reform in all three countries. During the years preceding tax reform in New Zealand and Australia tax rates were constant for both these countries. During 1987-88 rates initially went up at the corporate level
and down for individuals. Australian and New Zealand corporate and individual rates decreased in 1989, but went down relatively farther in New Zealand. Corporate rates in New Zealand went from 48 percent to 28 percent in 1989 (33 percent for 1990-92), while individual rates settled to 33 percent by 1990, as well. Corporate rates went from 49 percent to 39 percent in Australia in 1989, while individual tax rates did not go down until 1990 to 48 percent, and then to 47 percent for 1991-92. In Canada, corporate rates increased in 1972 and 1973, then dropped below pre-reform levels in 1974. Canadian individual rates dropped during the years surrounding adoption of the tax reform, then rebounded partially in 1977.
Descriptive statistics, means and medians, are given in Table 3 for the dependent and independent variables used in the regression model. The dependent investment measure is similar in all three countries. The DNOL variable is similar in New Zealand and Australia, slightly above 0.5 indicating that the majority of firms had an operating loss in the past five years. This variable is less than 0.5 in Canada. The cash flow ratio is much higher in Australia (0.48) than in either New Zealand (0.11) or Canada (0.22), although this appears to be skewed as the medians are similar across countries. Book-to-market and debt-to-equity ratios are higher in New Zealand than in Australia or Canada. The capital intensity and size variables are similar across countries, although Canada has slightly smaller firms on average. In Australia, 53 percent of the observations are after the tax reform act in 1988; in New Zealand, 58 percent of the observations are after tax reform; and in Canada 60 percent of the observations are after tax reform.
Table 4 provides Pearson correlation coefficients of the dependent and independent variables. Of particular note is the high negative correlation between the tax reform dummy variable and the statutory tax rate variable in each of the countries. This high collinearity makes it harder to find results on our specific variable of interest, TREF, which measures the effect of the tax reform acts in our study countries. However, if we did not include the tax rate variable we could not distinguish between the effects of the tax reforms and the effects of the statutory tax rate changes. Other variables with significant partial correlations include capital intensity (CINT) with size and lagged investment. To the extent there is multicollinearity it should not bias the coefficients, but makes it less likely to reject the null hypothesis of no significant relation with a firm’s investment.
Research Methodology
Regression analysis is used to test the effects of each country’s tax reform package on investment. The major tax change that is consistent across all three countries, implementation of a dividend imputation plan, was implemented in 1972 in Canada, 1988 in Australia and during the second half of 1987 in New Zealand. These countries are fairly homogeneous in their culture, tax and accounting systems.
The basic research model is the following:
(5)
where,
INVit = measure of investment for firm i, at time t.
TREFt = dummy measure of whether the tax reform has become effective.
RATEt = the maximum of either the statutory corporate tax rate or the individual tax rate at time t.
DNOLit = measure of the likelihood of a net operating loss for firm i, at time t. If the
firm has had a net operating loss in the last five years DNOL = 1, 0
otherwise.
BKMKTit = book-to-market ratio for firm i, at time t.
DEit = debt-to-equity ratio for firm i, at time t.
CFit = cash flow ratio for firm i, at time t.
PREDPit = interactive TREF dummy measure with the dividend payout ratio (DP) for
firm i, at time t: (1-TREF)*DP. This measure yields the pre-tax reform
coefficient.
PRECINTit = interactive TREF dummy measure with the capital intensity ratio (CINT)
for firm i, at time t: (1-TREF)*CINT. This measure yields the pre-tax
reform coefficient.
POSTDPit = interactive TREF dummy measure with the dividend payout ratio (DP) for
firm i, at time t: TREF*DP. This measure yields the post-tax reform
coefficient.
POSTCINTit = interactive TREF dummy measure with the capital intensity ratio (CINT)
for firm i, at time t: TREF)*CINT. This measure yields the post-tax
reform coefficient.
SIZEit = a measure of the size of firm i, at time t.
LINVit = lagged measure of investment for firm i, at time t-1.
This regression equation is estimated for each of the three countries to analyze effects of different tax changes. The coefficient on TREF is expected to be significant if tax reform affects investment, but may vary across country.
RESULTS
Regression Results by Country
Regression results for each country are reported in Table 5. By testing and comparing tax reforms in the three countries, we provide evidence on the incremental and combined effects of various reforms on corporate fixed investment.
New Zealand
In New Zealand, where the tax reform consisted primarily of the adoption of dividend imputation,[17] the positive relation which we observe between the TREF variable and corporate investment supports our hypothesis, indicating that dividend imputation stimulated corporate fixed investment in New Zealand. The other major tax change in New Zealand was a lowering of statutory tax rates. Tax rates, indicated by the variable RATE, were insignificant indicating that tax rates did not impact corporate investment. The negative relation between DNOL and corporate investment indicates that a firm’s earnings status impacts its level of capital investment. This negative relation could be due to the lack of benefits provided by dividend imputation to corporations with low marginal tax rates.
The variable PREDP, which measures the pre-imputation dividend payout ratio, was significant in New Zealand indicating that dividend payout ratio impacted corporate investment.
After dividend imputation dividend payout ratio was no longer associated with corporate investment as the coefficient on the variable POSTDP was not significant. Moreover, the F-value testing this change (POSTDP-PREDP=0) is insignificant at 2.55. Thus, we cannot conclude that dividend imputation significantly impacted this relation. PRECINT, which measures the pre-imputation level of capital intensity, was significant at the 0.01 level indicating that the level of capital intensity influenced corporate investment in New Zealand. Conversely, after dividend-imputation, corporate investment is no longer influenced by the level of capital intensity as the coefficient on POSTCINT is insignificant. The F-value of 7.50 obtained from testing whether the coefficient on POSTCINT is significantly different from the coefficient on PRECINT (POSTCINT-PRECINT=0) is significant at the 0.01 level. Thus, the tax reform significantly changed the relation between capital intensity and capital investment.
Australia
In Australia, the coefficient on TREF is again positive (0.67) and is significant at the 0.01 level, supporting our hypothesis that Australia’s tax reform, consisting of dividend imputation and the adoption of a capital gains tax, stimulated corporate capital investment. The variables DNOL and RATE are insignificant indicating no relation between NOL status and corporate tax rates and capital investment by Australian corporations.
For Australia, dividend payout ratios were not associated with corporate investment either before or after tax reform as the coefficients on both PREDP and POSTDP were insignificant. On the other hand, capital intensity influenced corporate investment before as well as after tax reform as the coefficients on PRECINT and POSTCINT were both positive and significant at the 0.01 level. However, the coefficient on POSTCINT is smaller than the coefficient on PRECINT. The F-value testing whether this difference is significantly different from zero (POSTCINT-PRECINT=0) is 23.38 and significant at the 0.01 level. Consistent with our New Zealand results, this suggests that capital intensity does not influence corporate investment as much after the tax reform as it did prior to the tax reform.
Canada
As in Australia, the Canadian tax reform package included dividend imputation and the imposition of capital gains taxes, as well as an investment tax credit directly targeting capital investment. As shown in table 5, the tax reform variable (TREF) is significant at the 0.10 level, indicating that its passage had an impact on corporate capital investment in Canada. In addition, the coefficient on RATE is positive (0.004) and significant at the 0.01 level indicating that the higher the corporate tax rate the greater the level of corporate investment. The coefficient on DNOL was not significant for the Canadian sample.
In Canada, dividend payout ratios impacted corporate investment both before and after tax reform as the coefficients on PREDP and POSTDP were significant at the 0.01 level and 0.05 level, respectively. However, the tax reform significantly affected the magnitude of the coefficients on dividend payout, indicating that the higher the dividend payout after tax reform the lower is the firm’s capital investment. The F-value of the test in difference of the coefficients on POSTDP and PREDP is 2.77 and is significant at the 0.10 level. The level of capital intensity also influenced corporate investment both before tax reform and after tax reform as the coefficients on PRECINT and POSTCINT were both significant at the 0.01 level. The F-Value testing whether there is a significant difference between the coefficients on POSTCINT and PRECINT is 0.84 and insignificant. Thus, we can conclude that the Canadian tax reform did not impact how capital intensity influenced corporate investment.
Control Variables
With regards to the control variables, the variable CFO, which measures cash flow ratio, had a significant positive impact on investment in all three countries while lagged capital investment (LINV) was significantly and positively related to current capital investment in Canada and New Zealand. Finally, the coefficients on the variables BKMKT and DE were insignificant for all three countries while the coefficient on SIZE was negative and significant in Canada and insignificant in the other two countries.
Portfolio Regression Results
We divide each of the three country samples into four portfolios based on dividend payout policies and capital intensity. Results in Table 6 provide evidence that New Zealand firms differed in their response to the tax legislation based on their historical dividend payout ratio and capital intensity. For Portfolio IV, high dividend payout and high capital intensity firms, the tax act variable (TREF) is significant and positive at the 0.10 level while for the other three portfolios TREF is not significant. This finding supports our hypothesis; the coefficient on PREDIV is positive and significant indicating that before tax reform, dividend policy influenced corporate investment in New Zealand whereas after tax reform, dividend policy no longer impacted investment as the variable POSTDIV was insignificant. The F-value to test whether the coefficients on POSTDIV and PREDIV are equal is 4.7 and is significant which indicates that post tax reform firms invest for economic reasons regardless of dividend policy. In contrast to the overall results reported in Table 4, the effect of capital intensity on investment was unaltered by tax reform in any of the four portfolios.
In Australia, a significant relationship between TREF and corporate investment indicates that the tax reform was successful in stimulating investment in fixed assets. Since Australia also adopted capital gains taxes for the first time, division of the Australian sample into high and low dividend portfolios enables us to disaggregate the effects of dividend imputation and capital gains taxes. Thus, by utilizing portfolios and variables measuring tax rates, we can determine the separate and combined effects of lower tax rates, dividend imputation and capital gains on fixed investment.
Results from Table 7 provide evidence tax reform impacted Australian firms differently. The coefficient on TREF was positive and significant for Portfolio III, while it was insignificant for the other three portfolios. These findings suggest that firms in Portfolio III were driving the results reported in Table 4. However, the predicted sign on TREF for Portfolio III was negative. A possible explanation for this finding is that the median dividend payout ratio for Australian firms (0.46) is still relatively high compared to the other two countries. For Portfolio I, tax reform changed how capital intensity influenced investment as the coefficient on POSTCINT is positive and significant. The F-value testing whether POSTCINT equals PRECINT is 3.66 and is significant at the 0.05 level and indicates that tax reform significantly altered how capital intensity impacted investment for firms in Portfolio I.
For firms in Portfolio II, dividend policy did not influence investment either before tax reform or after tax reform as the coefficients on PREDP and POSTDP were both insignificant. However, tax reform did significantly change how dividend policy influenced investment as the F-value from testing POSTDP equals PREDP is 11.74 and significant at the 0.01 level. This indicates that after tax reform, dividend policy has less influence on investment than before tax reform. Capital intensity impacted investment both before and after tax reform, as the coefficients on PRECINT and POSTCINT were both positive and significant at the 0.05 level. In addition, there was no change in how capital intensity impacted investment as the F-value testing whether PRECINT equals POSTCINT was 0.12 and not significant.
For Australian firms in Portfolio III, dividend policy did not impact investment either before or after tax reform as the coefficients on PREDIV and POSTDIV were insignificant. However, tax reform in Australia altered how capital intensity influenced investment as the coefficient on PRECINT was positive and significant at the 0.01 level while the coefficient on POSTCINT was insignificant. In addition, the F-value which test whether PRECINT equals POSTCINT was 18.25 and significant at 0.01 level indicating that tax reform in Australia significantly impacted how capital intensity influenced investment.
Finally, for Australian firms in Portfolio IV dividend policy and capital intensity did not impact investment either before tax reform or after as the coefficients on PREDIV, POSTDIV, PRECINT and POSTCINT were all insignificant.
Table 8 provides the results for the portfolio regressions for Canada. The results indicate that tax reform in Canada impacted the four Portfolios differently. For Portfolio IV, the coefficient on TREF had the predicted sign and was positive and significant at the 0.05 level. This indicates that tax reform impacted the capital investment of high dividend payout/high capital intensity firms and suggests dividend imputation and the ITC outweighed the increase in the capital gains tax. The coefficient on TREF for firms in Portfolio II, and III was insignificant while it was positive and significant for Portfolio I. However, the predicted sign for Portfolio I was negative. A possible explanation for this finding is that these firms are growing firms that expect to have a greater need for capital assets in the future. Thus, they used the advantages of tax reform to make major capital investments.
In Portfolio I, capital intensity had an impact on investment before tax reform as the coefficient on PRECINT was positive and significant while after tax reform, this was no longer the case as POSTCINT was insignificant. However, the F-Value of 2.47 to test whether there was a difference between POSTCINT and PRECINT was not significant. Also, for Portfolio 1 dividend policy did not influence investment either before or after tax reform.
For Portfolio II, dividend policy impacted investment before tax reform as the coefficient on PREDIV was significant whereas the insignificant coefficient on POSTDIV indicates that after tax reform, dividend policy no longer influenced investment. Conversely, capital intensity did not impact investment before tax reform as the coefficient on PRECINT was not significant while it did impact investment after tax reform as the coefficient on POSTCINT was positive and significant. However, for both dividend policy and capital intensity, the F-values to test whether POSTDIV equals PREDIV and POSCINT equals PRECINT were both insignificant indicating that tax reform did not significantly alter the influence dividend policy and capital intensity had on investment.
For Portfolio III, dividend policy had no impact on investment either before or after tax reform. On the other hand, capital intensity impacted investment both before and after tax reform as the coefficients on PRECINT and POSTCINT were both positive and significant. In addition, the coefficients on PRECINT and POSTCINT were not significantly different as the F-value to test whether there was a difference between the two was not significant.
For Portfolio IV, the coefficient on PREDIV was not significant and indicates that dividend policy had no impact on investment before tax reform while after tax reform, the coefficient on POSTDIV is negative and significant. The F-value to test whether PREDIV equals POSTDIV is 4.92 and significant at the 0.05 level. This indicates that tax reform altered how dividend policy influences the investment decisions of firms in Portfolio IV. The positive and significant coefficients on PRECINT and POSTCINT indicate that capital intensity had an impact on investment both before and after tax reform. In addition, tax reform did not significantly change in how capital intensity impacted investment as the F-value to test whether PRECINT equals POSTCINT was not significant.[18]
SUMMARY AND CONCLUDING REMARKS
The potential misallocation of economic resources has historically been a major concern about the efficiency of traditional corporate tax systems. For example Harberger (1962) demonstrates that the traditional double tax on corporate profits causes capital to flow out of the corporate sector, artificially reducing capital investment by corporations. This bias against capital investment in the corporate sector in turn results in inefficient pricing and consumption of corporate products. Continuing research indicates that these distortions can have significant macroeconomic economic effects, including lower growth of GNP.[19] An overriding objective of dividend imputation, the common tax reform enacted by New Zealand, Australia and Canada and examined in this study, was to reduce biases against the corporate form of business believed to be induced by traditional double tax systems.
By determining the impact of tax reforms on capital investment, policy-makers in the U.S. and elsewhere can make more informed decisions regarding tax policy as it affects capital investment. This study determines that dividend imputation resulted in increased corporate capital investment in all three countries. Moreover, in Australia it appears that the stimulus to corporate investment provided by dividend imputation overshadowed the possible negative effects of new capital gains taxes. In Canada, where the tax reform also resulted in enhanced capital investment, it is more difficult to disentangle the beneficial effects of dividend imputation and the new ITC. However, the results indicate that tax reform packages that combine various attributes including dividend imputation resulted in increased corporate capital investment.
Our tests of portfolios are based on the assumption that each portfolio is composed of firms which are completely unaffected by some aspect of its country’s tax reform. Obviously, relatively few corporations will remain unaffected by tax reforms of this magnitude. Thus, while only tentative conclusions can be drawn from our portfolio tests, the results support our primary findings. Furthermore, the results from the portfolios clearly indicate that corporations respond differently to tax reforms based on firm specific characteristics. Tax policy makers need to be aware that tax reforms may result not only in increased or decreased capital investment overall, but in a shifting of investment, with wealth effects, from one sector of the economy to another.
This possible shifting of investment raises one limitation in the interpretation of our results. The results indicate that capital investment by relatively large corporations was enhanced through tax reform. However, data limitations restrict our ability to measure capital investment within other sectors of the economy, and to control for non-tax influences on investment within those sectors. Thus, part of the increased investment that we document could represent a shift of investment from another economic sector rather than an outright increase.
Another limitation, common to all such studies, is that major tax reforms rarely represent discrete events. Most major tax reforms are composed of numerous changes with potentially offsetting effects. Moreover, time lags in the form of political debates and actual implementation of enacted changes raise many timing and measurement issues. In response to these issues, we have explored reasonable alternatives in how and when several variables are measured. The consistency of our results indicate that these issues do not materially affect the reliability of our findings, nor conclusions based upon them.
APPENDIX
This appendix outlines the basic mechanics of a typical dividend imputation plan and provides details of the tax reforms studied in this paper. The objective of a dividend imputation scheme is to levy only one level of tax on corporate income. It should be noted that this objective is already achieved in the taxation of traditional “pass through” business entities such as subchapter S corporations and partnerships. However, dividend imputation differs fundamentally from those taxation schemes in that under dividend imputation the tax is levied on the corporation, using corporate tax rates. At the corporate level a dividend imputation plan is quite similar to the traditional double tax system currently used in the U.S. Thus, it is the taxation of dividend distributions that sets dividend imputation plans apart from the traditional corporate tax system. An example of a typical dividend distribution under an imputation system clearly outlines these differences.
Assume a corporation earns $100 of taxable income and that the corporate tax rate is 35 percent. The corporation will pay $35 of tax on account of this income, leaving $65 available for distribution to shareholders. Assume now that the corporation declares and pays all $65 as a dividend to a noncorporate shareholder. That shareholder must include the amount of the dividend in his taxable income. The amount that the shareholder includes, however, is “grossed up” for the amount of tax that has been paid by the corporation and is attributable to the dividend. Thus, the shareholder includes in taxable income the $65 cash distribution as well as the $35 of taxes that the corporation paid.
At this point, it appears that the shareholder is worse off under dividend imputation than with a traditional double tax system since his taxable income is $35 higher. However, the tax relief takes the form of a tax credit that the shareholder now receives. The amount of the credit is equal to the amount of tax that the corporation has paid on the grossed up dividend, $35 in this example. Assuming that individual and corporate tax rates are identical at 35 percent, the shareholder will have a gross tax increase of $35 (35 percent of the grossed-up dividend) and a perfectly offsetting tax credit of $35. In summary, the cash dividend has no net tax effect on the shareholder and the income has been subjected to only one level of tax.
This example can be modified to illustrate the outcome of dividend imputation when individual and corporate rates vary. Assume that the individual tax rate in the example above is 40 percent. As before, the shareholder will have to report grossed-up dividend income of $100. However, the $35 imputation credit for taxes paid by the corporation will not entirely offset the $40 of individual taxes due. As can be seen, an imputation system where individual tax rates exceed corporate tax rates results in only partial elimination of double taxation.
In cases where the corporate tax rate exceeds individual tax rates, the results can be more complex. Moreover, since most tax systems have graduated individual tax rates, this situation can occur at almost any time even if maximum individual rates exceed maximum corporate rates. Continuing the original example where a corporation distributes a $65 cash dividend, now assume that the recipient is an individual shareholder in a 20 percent marginal tax bracket. The shareholder includes the grossed-up dividend of $100 in taxable income, which results in an increased tax due of $20. However, the dividend carries a $35 imputation credit which, if fully allowed, will result in a total net tax of just $20, or 20 percent, on the original corporate income. An individual shareholder under these circumstances would actually reduce his tax liability whenever he receives a cash dividend.
In order to insure that all corporate income is subject to at least one level of tax, corporations generally maintain a cumulative record of taxes paid. The balance of this account represents the amount of tax credits available on future dividend distributions and is increased when corporate income taxes are paid and reduced whenever the corporation makes a dividend distribution. Once the account balance is zero, any additional dividend distributions will not carry an imputation credit.
Under certain imputation systems, one possible result of this is that corporations may find less incentive to make use of traditional tax shields. By using a tax shield, the corporation can successfully reduce its corporate tax. However, this strategy also reduces the imputation credits available to shareholders. If the corporation makes distributions in excess of taxable income, shareholders will have to include the grossed-up dividend in their taxable income but receive no offsetting imputation credit. In summary, an additional tax shield for a high-dividend corporation may simply “shift” the tax burden from the corporation to the shareholders.
Tax Reform in Australia
The dividend imputation plan enacted by Australia became effective on July 1, 1987, and closely resembles the generic plan described above. It is important to note that excess imputation credits are not refundable to any shareholder, including tax exempt shareholders. This is not a major factor in Australia for superannuation funds since such funds have been subject to a 15 percent tax rate since 1988. However, imputation credits on dividends paid to tax exempt shareholders or shareholders with relatively low taxable income might be lost.
In addition to dividend imputation, the Australian government imposed a capital gains tax on assets acquired after September 19, 1995.[20] This had the effect of grandfathering not only unrealized gains as of that date, but any additional gains on assets already held by shareholders on September 19, 1995. Recognized capital gains receive no preference in the form of lower rates in Australia. However, when determining the amount of gain to be recognized on disposition of a capital asset, the owner’s basis is indexed for inflation.
Tax Reform in Canada
Canada’s dividend imputation scheme, which became effective in January of 1972, differs from that of most countries in that there is no direct link between taxes paid at the corporate level and the generation of imputation credits to shareholders. Shareholders receive a credit based on grossed-up dividends regardless of the amount of taxes actually paid, or not paid, by the distributing corporation. One effect of this is that traditional tax shields which reduce corporate taxes will not limit dividend credits as they can in New Zealand and Australia. Canadian corporations are also relieved of the burden of tracking the amount of imputation credits available to shareholders. As with Australia and New Zealand, credits in excess of pre-credit tax liability are not refundable.
Canada’s tax reform also imposed a capital gains tax on stock held for investment. One-half of realized gains were taxes at ordinary income tax rates. This is similar to Australia’s system, except there was no grandfathering of pre-existing gains and no indexing for inflation. Finally, Canada instituted an ITC for investment in non-real estate business assets.
References
Amoako-Adu, B. The Canadian Tax Reform and Its Effect on Stock Prices: A Note. Journal of Finance (December 1983), 1669-1675.
Auerbach, A.J. 1983. Taxation, Corporate Financial Policy and the Cost of Capital. Journal of Economic Literature 21: 905-940.
_______., and K.A. Hassett. 1991.Recent U.S. Investment Behavior and the Tax Reform Act of 1986: A Disaggregate View. Carnegie-Rochester Conference Series on Public Policy 35: 185-215.
Ayres, Frances L. 1987. An Empirical Assessment of the Effects of the Investment Tax Credit Legislation on Returns to Equity Securities. Journal of Accounting and Public Policy 6: 115-137.
Bernanke, B.S., H. Bohn, M. Devereux and F. Schiantarelli. 1994. Alternative Non-Nested Specification Tests of Time-Series Investment Models. Journal of Econometrics 37:
1-38.
Cassou, S.P. 1997. The Link Between Tax Rates and Foreign Direct Investment. Applied Economics 29: 1295-1301.
Chirinko, R.S. 1986. Business Investment and Tax Policy: A Perspective on Existing Models and Empirical Results." National Tax Journal 39 (2): 137-155.
______,. 1993. Business Fixed Investment Spending: A Critical Survey of Modelling Strategies, Empirical Results, and Policy Implications. Journal of Economic Literature 31: 1875-1911.
Christie, A.A. 1990. Aggregation of Test Statistics: An Evaluation of the Evidence on Contracting and Size Hypotheses. Journal of Accounting and Economics 12: 15-36.
Chung, K., and C. Charoenwong. 1991. Investment options, assets in place, and the risk of stocks. Financial Management 10 (Autumn): 21-33.
Clark, P.K. 1979. Investment in the 1970s: Theory, Performance, and Prediction. Brookings Papers on Economic Activity 1: 73-113.
_______,.1993. Tax Incentives and Equipment Investment. Brookings Papers on Economic Activity 1: 317-339.
Collins, D.W., and S. P. Kothari. 1989. An analysis of the intertemporal and cross-sectional determinants of the earnings response coefficients. Journal of Accounting and Economics 9: 111-138.
Courtenay, Stephen M., Robert P. Crum, and Stuart B. Keller. 1989. Differential Reactions to Legislative Signaling During the Enactment of ERTA and TEFRA: An Empirical Investigation of Market Returns and Volume. Journal of Accounting and Public Policy 8: 283-321.
Cummins, J.G. and K.A. Hassett. 1992. The Effects of Taxation on Investment: New Evidence From Firm Level Panel Data. National Tax Journal 45(3): 243-251.
_______., K.A. Hassett, and R.G. Hubbard. 1994. A Reconsideration of Investment Behavior Using Tax Reforms as Natural Experiments. Brookings Papers on Economic Activity 2: 1-74.
_______., K.A. Hassett, and R.G. Hubbard. 1996. Tax Reforms and Investment: A Cross-Country Comparison. Journal of Public Economics 62: 237-273.
Galeotti, M., F. Schiantarelli and F. Jaramillo. 1994. Investment Decisions and The Role of Debt, Liquid Assets and Cash Flow: Evidence from Italian Panel Data. Applied Financial Economics 4: 121-132.
Gilchrist, S. and C.P. Himmelberg. 1995. Evidence on The Role of Cash Flow For Investment. Journal of Monetary Economics 36: 541-572.
Gravelle, J. 1991. Corporate Tax Integration: Issues and Options. Washington, D.C.: Congressional Research Service.
Gravelle, J. and L.J. Kotlikoff. 1989. The Incidence and Efficiency Costs of Corporate Taxation When Corporate and Noncorporate Firms Produce the Same Goods. Journal of Political Economy 97: 749-790.
Hall, R.E., and D.W. Jorgenson. 1967. Tax Policy and Investment Behavior. American Economic Review 57: 391-414.
Harberger, A.C. 1962. The Incidence of the Corporate Income Tax. Journal of Political Economy 70: 215-240.
Head, J.G. 1993. Imputation in the context of taxation reform. Bureau of Industry Economics, Dividend Imputation: Policy Forum September 1993 (p. 9-69). Canberra: Australian Government Publishing Service.
Hubbard, R. G., A.K. Kashyap, and T.M. Whited. 1995. Internal Finance and Firm Investment. Journal of Money, Credit, and Banking 27(3): 683-701.
Jorgenson, D.W. 1963. Capital Theory and Investment Behavior. American Economic Review 53: 247-259.
Kern, Beth B. 1994. The Redistribution of Corporate Plant and Equipment as a Result of the Economic Recovery Act of 1981. Journal of Accounting and Public Policy 13: 225-252.
Kinney, Michael and R.H. Trezevant. 1993. Taxes and the Timing of Corporate Capital Expenditures. The Journal of the American Taxation Association 15 (Fall): 40-62.
Lewellen, W., C. Loderer, and K. Martin. 1987. Executive compensation and executive incentive problems: An empirical analysis. Journal of Accounting and Economics 9: 287-310.
Lev, Baruch and Theodore Sougiannis. 1996. The Capitalization, Amortization, and Value-Relevance of R&D. Journal of Accounting and Economics 21(1): 107-138.
Moon, Philip, and Stewart Hodges. 1989. Implications of the Recent Tax Changes For Corporate Capital Investment. Journal of Business Finance and Accounting 16(1): 25-39.
Moore, Michael L., B.M. Steece and C.W. Swenson. 1987. An Analysis of the Impact of State Income Tax Rates and Bases on Foreign Investment. The Accounting Review 4: 671-685.
Morgan, E.J. 1992. Investment Behaviour and Government Tax Policy: A Survey of The Largest UK Firms. Journal of Business Finance and Accounting 19(6): 789-815.
Oliner, S., G. Rudebusch and D. Sichel. 1995. New and Old Models of Investment: A Comparison of Forecasting Performance. Journal of Money, Credit and Banking 27: 806-826.
Omer, T., K. Molloy, and D. Zeibart. 1991. Measurement of Effective Corporate Tax Rates Using Financial Statement Information. Journal of the American Taxation Association 13(Spring): 57-72.
Rosacker, Robert E. and R.W. Metcalf. 1993. United States Federal Tax Policy Surrounding The Investment Tax Credit: An Empirical Investigation of Enactments, Rate Enhancement, and Repeals. Advances In Taxation 5: 219-246.
Schulman, C.T., D.W. Thomas, K.F. Sellers, and D.B. Kennedy. 1996. Effects of Tax Reform and Capital Gains on Corporate Leverage. National Tax Journal 1: 31-54.
Shapiro, M. 1986. Investment, Output and the Cost of Capital. Brookings Papers on Economic Activity 1: 111-152.
Smith, C.W. and R.L. Watts. 1992. The investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Financial Economics. 263-292.
Swenson, Charles W. 1989. Tax Regimes and the Demand for Risky Assets: Some Experimental Market Evidence. The Journal of the American Taxation Association (Fall): 54-76.
_______,. 1987. An Analysis of ACRS During Inflationary Periods. The Accounting Review 1: 117-136.
Tobin, J. 1969. A General Equilibrium Approach to Monetary Theory. Journal of Money, Credit, and Banking 1(1): 15-29.
Vines, Cynthia C., C.W. Swenson, and W. Rodgers. 1994. Testing the Determinants of State Business Tax Burdens. Journal of Accounting and Public Policy 13: 205-224.
Table 1
Expected Relation of Tax Reform with Capital Expenditures
for New Zealand, Australia, and Canada
Portfolios Partitioned by Dividend Payout and Capital Intensity
A: New Zealand
| |DIVIDEND PAYOUT |
| | |
|C I | |
|A N | |
|P T | |
|I E Low| |
|T N | |
|A S | |
|L I | |
|T High | |
|Y | |
| |Low |High |
| |Insignificant TREF |Significant Positive TREF Coefficient |
| |Coefficient | |
| | |II |
| |I | |
| |Significant Positive |Significant Positive |
| |TREF Coefficient |TREF Coefficient |
| | | |
| |III |IV |
B: Australia
| |DIVIDEND PAYOUT |
| | |
|C I | |
|A N | |
|P T | |
|I E Low| |
|T N | |
|A S | |
|L I | |
|T High | |
|Y | |
| | |
| |Low |High |
| |Insignificant TREF |Significant Positive TREF Coefficient |
| |Coefficient | |
| | |II |
| |I | |
| |Significant Negative TREF Coefficient|Significant Positive TREF Coefficient |
| | | |
| |III |IV |
C: Canada
| |DIVIDEND PAYOUT |
| | |
|C I | |
|A N | |
|P T | |
|I E Low| |
|T N | |
|A S | |
|L I | |
|T High | |
|Y | |
| |Low |High |
| |Significant Negative TREF |Significant Positive TREF Coefficient|
| |Coefficient | |
| | |II |
| |I | |
| | |Significant Positive TREF Coefficient|
| |? | |
| | |IV |
| |III | |
Table 2
Descriptive Data on Tax Rates in Australia and New Zealand
| |New Zealand |Australia |Canada |
| |Maximum |Maximum |Maximum |Maximum | |Maximum |Maximum |
|YEAR |Corporate |Individual |Corporate |Individual |YEAR |Corporate |Individual |
| |Tax Rate |Tax Rate |Tax Rate |Tax Rate | |Tax Rate |Tax Rate |
|1982 |45% |60% |46% |60% |1968 |51.4% |60% |
|1983 |45% |66% |46% |60% |1969 |51.4% |60% |
|1984 |45% |66% |46% |60% |1970 |46.7% |60% |
|1985 |45% |66% |46% |60% |1971 |46.5% |57.6% |
|1986 |45% |66% |46% |60% |1972 |49% |45.6% |
|1987 |48% |57% |49% |57% |1973 |50.6% |47% |
|1988 |48% |48% |49% |49% |1974 |48.2% |47% |
|1989 |28% |40.5% |39% |49% |1975 |46% |47% |
|1990 |33% |33% |39% |48% |1976 |46% |47% |
|1991 |33% |33% |39% |47% |1977 |46% |51.7% |
Table 3
Descriptive Statistics of Regression Variables
Means (Medians)
|Variable |New Zealand |Australia |Canada |
| | | | |
|Investment (INV) |0.15 |0.16 |0.15 |
| |(0.04) |(0.04) |(0.04) |
| | | | |
|Tax Reform Dummy |0.58 |0.53 |0.60 |
|(TREF) |(1.00) |(1.00) |(1.00) |
| | | | |
|Net Operating Loss Dummy (DNOL) |0.57 |0.56 |0.37 |
| |(1.00) |(1.00) |(0) |
| | | | |
|Statutory Tax Rate |0.49 |0.54 |0.53 |
|(RATE) |(0.48) |(0.49) |(0.51) |
| | | | |
|Book-to-Market Ratio |5.15 |2.23 |0.95 |
|(BKMKT) |(1.52) |(0.93) |(0.61) |
| | | | |
|Debt-to-Equity Ratio |1.78 |1.01 |1.48 |
|(DE) |(1.16) |(1.05) |(1.12) |
| | | | |
|Cash Flow Ratio |0.11 |0.48 |0.22 |
|(CF) |(0.07) |(0.10) |(0.11) |
| | | | |
|Dividend Payout Ratio |0.31 |0.21 |0.48 |
|(DP) |(0.42) |(0.46) |(0.28) |
| | | | |
|Capital Intensity Ratio |0.40 |0.43 |0.45 |
|(CINT) |(0.35) |(0.41) |(0.42) |
| | | | |
|SIZE |5.75 |5.85 |3.81 |
| |(5.57) |(5.81) |(4.07) |
|Lagged Investment |0.19 |0.17 |0.15 |
|(LINV) |(0.05) |(0.03) |(0.04) |
Definition of Variables:
INV = measure of investment for firm i, at time t.
TREF = dummy measure of whether tax reform is available for firm i, at time t.
DNOL = measure of the likelihood of a net operating loss for firm I, at time t. If the
firm has had a net operating loss in the last five years DNOL = 1, 0
otherwise.
RATE = the maximum of the statutory corporate tax rate or the individual tax rate
at time t.
BKMKT = book-to-market ratio for firm i, at time t.
DE = debt-to-equity ratio for firm i, at time t.
CF = cash flow ratio for firm i, at time t.
DP = dividend payout ratio for firm i, at time t.
CINT = capital intensity ratio for firm i, at time t.
SIZE = a measure of the size of firm i, at time t.
LINV = lagged measure of investment for firm i, at time t-1.
Table 4
Correlation Analysis
A: New Zealand – Pearson Correlation Coefficients
| |INV |TREF |DNOL |RATE |BKMKT |DE |CF |DP |CINT |SIZE |
|TREF |-0.221 | | | | | | | | | |
|DNOL |0.097 |-0.108 | | | | | | | | |
|RATE |0.227 |-0.947 |0.081 | | | | | | | |
|BKMKT |0.006 |-0.279 |0.012 |0.276 | | | | | | |
|DE |0.491 |-0.110 |0.075 |0.096 |-0.056 | | | | | |
|CF |0.202 |0.073 |0.121 |-0.060 |-0.010 |0.033 | | | | |
|DP |0.056 |-0.031 |0.064 |-0.018 |0.016 |0.004 |0.031 | | | |
|CINT |0.399 |-0.027 |0.179 |0.014 |0.161 |0.214 |0.008 |0.055 | | |
|SIZE |0.174 |0.037 |-0.228 |-0.038 |0.060 |0.174 |0.097 |0.035 |0.307 | |
|LINV |0.639 |-0.187 |0.146 |0.211 |0.010 |0.637 |0.159 |0.029 |0.456 |0.192 |
B: Australia – Pearson Correlation Coefficients
| |INV |TREF |DNOL |RATE |BKMKT |DE |CF |DP |CINT |SIZE |
|TREF |0.075 | | | | | | | | | |
|DNOL |-0.229 |-0.195 | | | | | | | | |
|RATE |-0.094 |-0.984 |0.172 | | | | | | | |
|BKMKT |-0.094 |-0.023 |0.075 |0.030 | | | | | | |
|DE |-0.003 |-0.037 |-0.002 |0.023 |0.044 | | | | | |
|CF |0.38 |-0.062 |0.054 |0.067 |-0.035 |-0.006 | | | | |
|DP |0.002 |-0.019 |-0.031 |0.018 |0.012 |-0.012 |-0.001 | | | |
|CINT |0.150 |0.012 |0.107 |-0.026 |0.067 |-0.058 |0.041 |0.029 | | |
|SIZE |0.131 |0.166 |-0.273 |-0.179 |0.239 |0.043 |-0.161 |0.010 |-0.041 | |
|LINV |0.168 |0.127 |-0.110 |-0.148 |-0.042 |0.027 |0.070 |0.002 |0.094 |0.125 |
C: Canada – Pearson Correlation Coefficients
| |INV |TREF |DNOL |RATE |BKMKT |DE |CF |DP |CINT |SIZE |
|TREF |0.043 | | | | | | | | | |
|DNOL |-0.019 |-0.228 | | | | | | | | |
|RATE |-0.060 |-0.787 |0.264 | | | | | | | |
|BKMKT |-0.063 |0.103 |-0.261 |-0.084 | | | | | | |
|DE |0.102 |0.061 |0.038 |-0.042 |-0.062 | | | | | |
|CF |0.175 |-0.024 |-0.005 |0.022 |0.020 |-0.038 | | | | |
|DP |-0.017 |-0.022 |0.029 |0.019 |-0.016 |-0.013 |0.334 | | | |
|CINT |0.471 |-0.063 |0.016 |0.061 |-0.076 |0.161 |-0.025 |-0.026 | | |
|SIZE |0.308 |0.127 |-0.382 |-0.126 |0.158 |-0.0003 |0.007 |0.001 |-0.177 | |
|LINV |0.571 |0.062 |0.073 |-0.092 |-0.052 |0.103 |0.147 |-0.014 |0.468 |-0.300 |
Bold indicates significant p-value < 0.05
Table 5
Regression Results by Country
(5)
| |New Zealand |Australia |Canada |
| |N = 158 |N = 1,010 |N = 1,776 |
|Independent Variable |Expected Sign |Parameter Estimate |Parameter Estimate |Parameter Estimate |
| | | | | |
|Intercept |? |-0.61 |-2.45*** |-0.21 |
| | | | | |
|TREF |+ |0.46** |0.67* |0.04*** |
| | | | | |
|RATE |? |0.003 |.034 |0.004* |
| | | | | |
|DNOL |- |-0.11*** |0.01 |-0.002 |
| | | | | |
|BKMKT |- |0.004 |-0.01 |0.02 |
| | | | | |
|DE |? |0.03 |-0.01 |0.003 |
| | | | | |
|CF |+ |0.45* |0.002* |0.00004* |
| | | | | |
|PREDP |- |0.16** |-0.01 |-0.0001* |
| | | | | |
|PRECINT |+ |1.21* |1.87* |0.43* |
| | | | | |
|POSTDP |- |0.006 |0.0001 |-0.02** |
| | | | | |
|POSTCINT |+ |-0.03 |0.68* |0.39* |
| | | | | |
|SIZE |? |0.003 |-0.001 |-0.02* |
| | | | | |
|LINV |+ |0.31* |0.01 |0.34* |
| | | | |
|Model F-Value |13.57* |27.94* |140.99* |
| | | | |
|Adjusted R-square |0.49 |0.24 |0.49 |
|TEST: | | | |
|PREDP-POSTDP = 0 |2.55 |0.68 |2.77*** |
|F-Value | | | |
|TEST: | | | |
|PRECINT – POSTCINT = 0 |7.50* |23.38* |0.84 |
|F-Value | | | |
*indicates significance at 0.01 level.
**indicates significance at 0.05 level.
***indicates significance at 0.10 level.
Table 6
New Zealand Portfolio Regression Results
(5)
| |I |II |III |IV |
|Independent |Low DP |HighDP |Low DP |High DP |
|Variables |Low CINT |Low CINT |High CINT |High CINT |
| |N = 42 |N = 38 |N = 29 |N = 49 |
| | | | | |
|Intercept |-0.04 |0.02 |-1.30 |-1.63 |
| | | | | |
|TREF |-0.01 |-0.11 |0.86 |1.37*** |
| | | | | |
|RATE |0.001 |0.001 |0.01 |0.01 |
| | | | | |
|DNOL |-0.06*** |0.04 |-0.06 |-0.05 |
| | | | | |
|BKMKT |0.0004 |-0.001 |0.01 |0.01 |
| | | | | |
|DE |0.01 |0.01 |0.03 |-0.06 |
| | | | | |
|CF |0.09 |-0.02 |0.88** |2.35* |
| | | | | |
|PREDP |-0.03 |-0.04 |-0.17 |0.50** |
| | | | | |
|PRECINT |0.02 |-0.44 |1.81 |0.65 |
| | | | | |
|POSTDP |0.004 |-0.07 |0.13 |-0.18 |
| | | | | |
|POSTCINT |-0.14 |0.06 |-0.20 |-0.59 |
| | | | | |
|SIZE |0.01 |0.02 |0.01 |0.02 |
| | | | | |
|LINV |-0.21 |0.23 |0.13 |0.48 |
| | | | | |
|Model F-Value |1.01 |0.96 |1.84 |5.85* |
| | | | | |
|Adjusted R-square |0.003 |0.00 |0.27 |0.55 |
|TEST: | | | | |
|PREDP-POSTDP = 0 |0.48 |0.03 |0.34 |4.70** |
|F-Value | | | | |
|TEST: | | | | |
|PRECINT – POSTCINT = 0 |0.06 |0.23 |1.69 |0.52 |
|F-Value | | | | |
*indicates significance at 0.01 level.
**indicates significance at 0.05 level.
***indicates significance at 0.10 level.
Table 7
Australia Portfolio Regression Results
(5)
| |I |II |III |IV |
|Independent |Low DP |HighDP |Low DP |High DP |
|Variables |Low CINT |Low CINT |High CINT |High CINT |
| |N = 244 |N = 256 |N =244 |N = 265 |
| | | | | |
|Intercept |-2.05 |-0.07 |-8.54 |0.16 |
| | | | | |
|TREF |-0.07 |0.07 |4.29* |-0.20 |
| | | | | |
|RATE |0.03 |-0.001 |10.88 |-0.14 |
| | | | | |
|DNOL |-0.03 |-0.003 |-0.29 |0.08** |
| | | | | |
|BKMKT |0.0004 |0.01 |-0.02 |-0.01 |
| | | | | |
|DE |-0.002 |0.05** |-0.07 |0.02 |
| | | | | |
|CF |0.001 |0.06*** |0.002* |0.62* |
| | | | | |
|PREDP |-0.0002 |-0.004 |-0.87 |-0.002 |
| | | | | |
|PRECINT |0.34 |0.33*** |6.04* |-0.16 |
| | | | | |
|POSTDP |0.0001 |-0.16 |-0.001 |-0.01 |
| | | | | |
|POSTCINT |1.39* |0.44** |-0.06 |0.01 |
| | | | | |
|SIZE |0.04** |0.003 |-0.04 |0.01 |
| | | | | |
|LINV |-0.02 |0.01 |-0.06 |0.06 |
| | | | | |
|Model F-Value |3.06* |5.72* |9.03* |7.33* |
| | | | | |
|Adjusted R-square |0.09 |0.18 |0.28 |0.22 |
|TEST: | | | | |
|PREDP-POSTDP = 0 |0.003 |11.74* |1.58 |0.09 |
|F-Value | | | | |
|TEST: | | | | |
|PRECINT – POSTCINT = 0 |3.66** |0.12 |18.26* |0.21 |
|F-Value | | | | |
*indicates significance at 0.01 level.
**indicates significance at 0.05 level.
***indicates significance at 0.10 level.
Table 8
Canada Portfolio Regression Results
(5)
| |I |II |III |IV |
|Independent |Low DP |HighDP |Low DP |High DP |
|Variables |Low CINT |Low CINT |High CINT |High CINT |
| |N = 509 |N = 376 |N =411 |N = 480 |
| | | | | |
|Intercept |-0.23 |0.07 |-0.42 |-0.35* |
| | | | | |
|TREF |0.07*** |-0.05 |0.14 |0.13** |
| | | | | |
|RATE |0.004** |0.001 |0.36 |0.004*** |
| | | | | |
|DNOL |0.02 |0.05 |-0.10* |0.11* |
| | | | | |
|BKMKT |0.02 |-0.11 |0.05 |-0.17*** |
| | | | | |
|DE |-0.003 |-0.01*** |-0.00001 |0.01*** |
| | | | | |
|CF |0.000005* |0.000003* |0.00001* |0.00004* |
| | | | | |
|PREDP |0.03*** |-0.0001* |-0.15 |0.023 |
| | | | | |
|PRECINT |0.31* |0.03 |0.76* |0.67* |
| | | | | |
|POSTDP |0.02 |0.004 |-0.31 |-0.05* |
| | | | | |
|POSTCINT |0.07 |0.17** |0.57* |0.59* |
| | | | | |
|SIZE |-0.01 |-0.0013 |-0.01 |-0.02* |
| | | | | |
|LINV |0.15* |0.33* |0.36* |0.25* |
| | | | | |
|Model F-Value |5.89* |14.98* |13.80* |102.77* |
| | | | | |
|Adjusted R-square |0.10 |0.31 |0.27 |0.72 |
|TEST: | | | | |
|PREDP-POSTDP = 0 |0.14 |0.06 |0.68 |4.92** |
|F-Value | | | | |
|TEST: | | | | |
|PRECINT – POSTCINT = 0 |2.47 |1.23 |0.77 |0.74 |
|F-Value | | | | |
*indicates significance at 0.01 level.
**indicates significance at 0.05 level.
***indicates significance at 0.10 level.
-----------------------
[1] See Shapiro (1986) for an excellent summary of this research.
[2] The various dividend imputation plans adopted by New Zealand, Australia and Canada do not necessarily eliminate all income taxes on dividends, but are designed to eliminate or sharply reduce double taxation and resulting biases against corporate distributions and investment.
[3] For reviews of these models and empirical studies linking taxes to investment see Auerbach (1983) and Chirinko, (1986; 1993). Chirinko (1993) concludes that the evidence of such a link is weak at best.
[4] See, for example, Clark (1979; 1993), Bernanke et al. (1988), and Oliner et al. (1995).
[5] Moore et al. (1987, p. 673) define a unitary tax system as one that requires filing a consolidated corporate income tax return that includes affiliates considered to be part of the unitary business. They indicate that most foreign firms are treated as either a worldwide or domestic combination. Under a worldwide combination all of the foreign firm’s affiliates, both domestic and foreign, are included in the tax base. The domestic combination requires just the affiliates in the United States to be included in the tax base.
[6] The 1980s were a period of low inflation for the United States. To analyze the impact of inflation on ACRS, Swenson (1987) conducted a Monte Carlo simulation.
[7] We provide an appendix that outlines the details of the tax reform acts in each of the three countries. See Appendix.
[8] For purposes of brevity, equation (2) reflects the cost of capital for an unlevered corporation. Aside from the change in corporate tax rates, the major tax legislation examined in this study was generally designed to directly impact equity rather than debt financing by adding or removing biases against equity financing. Indirect effects to the cost of debt financing, arising from a change in equilibrium debt to equity ratios for firms, may have occurred.
[9] Empirical evidence of this anticipated market adjustment is reported by Amoako-Adu (1983). He reports that high-dividend paying stocks increased significantly in value as a result of the Canadian tax reform.
[10] While Canada enacted all three of these changes, Australia two, and New Zealand only one, all three samples are partitioned the same to provide comparability.
[11] Prior research (Kinney and Trezevant 1993; Kern 1994) finds that various deflators have no impact on the results.
[12] Corporate and individual tax rates could impact the investment equation in a number of ways. Corporate taxes clearly decrease the net after-tax cash flows from an investment. However, most capital investments provide attractive tax shields which increase net cash flows. Individual rates, as illustrated in the development of our hypotheses, affect corporate cost of capital.
[13] Realizing that this is at best a proxy measure for low marginal tax rates, we also tested alternative classification schemes. Firms were alternatively classified as NOL firms if they had an NOL during the most recent year, a cumulative NOL during the past three years, or an NOL in the succeeding year. None of these alternative classification schemes changed the significance of the variables of interest.
[14] Lewellen, Loderer and Martin (1987) and Chung and Charoenwong (1991) also use this measure to control for growth opportunities.
[15] We do not use research intensity since R&D is incorporated in our measure of investment, the dependent variable.
[16] The capital gains tax went into effect in 1986. As a sensitivity test we also examine Australian data by removing this transition year, 1986. This did not qualitatively affect the results.
[17] The tax reform also resulted in reduced individual and corporate tax rates, but we control for this with the net operating loss dummy variable and the statutory tax rate variables. To test the sensitivity of these variables to our definition, we also test the model using dummy variables based on the most current year taxable income (negative = 1, 0 otherwise) and the next two years taxable income (negative =1, 0 otherwise) as a measure of net operating loss. These alternate forms did not affect the results; signs and significance of the test variables remained the same.
To test the sensitivity of the results to the measure of the statutory tax rate, we used both the maximum corporate and maximum individual rates, rather than the maximum of either rate, with no effect to our results. We also used an estimate of firm specific marginal tax rates (Kinney and Trezevant 1993), which did not impact the results on TREF. In addition, we added a gross national product (GDP) variable, however this did not affect the model. Tax rates vary by year and are also a proxy for the macroeconomic environment in any given year.
[18] Sensitivity to the macroeconomic environment is examined by including a gross domestic product (GDP) variable is added to equation (1) to control for effects to the economy in general. Adding this variable does little to the explanatory power of the model. The variables’ coefficients and signs are unaffected.
[19] A more detailed discussion of this area of research is beyond the scope of this paper. However, Gravelle (1991) and Gravelle and Kotlikoff (1989) provide excellent summaries and discussions of the ongoing research into the macroeconomic effects of corporate taxation.
[20] Although dividend imputation and the capital gains tax had different effective dates, they were clearly part of a integrated package of tax reforms. Head (1993) states that “imputation … did in fact form part of a substantial package of income tax reforms which included the introduction of a new indexed realisations tax on capital gains and other base-broadening measures..”
-----------------------
[pic]
[pic]
[pic]
[pic]
[pic]
[pic]
[pic]
[pic]
[pic]
[pic]
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
Related searches
- capital investment decisions pdf
- introduction to capital investment decisions
- capital investment decisions examples
- importance of capital investment decisions
- capital investment decision analysis
- capital investment decision making
- making capital investment decision
- why is capital investment important
- capital investment decision techniques
- capital investment analysis template
- capital investment roi calculator
- blackrock capital investment corporation