- Community
- Research
- Forms, Rates & Tables
- Government
- Executive
- Judicial
- Legislative
- Treasury
- Comments
- Continuing Professional Education
- Determination Letter
- Email Chief Counsel Advice
- Exempt Organizations Update
- Field Service Advice
- Forms and Instructions
- General Counsel Memoranda
- IRS Announcements
- IRS Fact Sheet
- IRS Forms
- IRS Legal Memoranda
- IRS Notices
- Information Release
- Internal Revenue Code
- Letter Rulings
- Publications
- Regulations
- Revenue Procedures
- Revenue Rulings
- Technical Advice Memoranda
- Treasury Decisions

- Management
- Marketing
- Planning

- News
- Articles
- Case Studies
- Technical Reports

# Study of Intergenerational Transfers

## Summary

**PGDC SUMMARY:**

The Treasury Department authored a paper in February 1996 which "investigates the effects of bequest taxes and the income of children on the lifetime charitable contributions of parents." The results of the study conclude, in part, that the greater the amount of the children's income, the greater the lifetime charitable contributions by the parents. In addition, with the caveat that the data used for an analysis of charitable bequests by parents in relation to their children's income is limited due to factors such as one parent is typically still alive, the authors' research suggests that "charitable bequests increase more than proportionally with the size of the estate" and that "children's income has a small influence on the charitable bequests of parents."

**FULL TEXT:**

CHARITABLE CONTRIBUTIONS AND INTERGENERATIONAL TRANSFERS /*/

by

GERALD AUTEN AND DAVID JOULFAIAN /**/

U.S. DEPARTMENT OF THE TREASURY

OTA Paper 72

February 1996

OTA Papers is an occasional series of reports on the research, models, and datasets developed to inform and improve Treasury's tax policy analysis. The papers are works in progress and subject to revision. Views and opinions expressed are those of the authors and do not necessarily represent official Treasury positions or policy. OTA Papers are distributed in order to document OTA analytic methods and data and invite discussion and suggestions for revision and improvement. Comments are welcome and should be directed to the authors.

Office of Tax Analysis

U.S. Department of the Treasury

Washington, DC 20220

gerald.auten@do.treas.gov

david.joulfaian@do.treas.gov

Charitable Contributions and Intergenerational Transfers

Abstract

This paper investigates the effects of bequest taxes and the income of children on the lifetime charitable contributions of parents. Using matched income tax records for parents and children, the results show a positive elasticity of 0.6 for contributions with respect to the tax price of bequests. The paper also finds that the income of children affects the amount that parents contribute to charity. The results show a positive elasticity of up to 0.14 for contributions by parents with respect to the income of children, implying that when children are better off, parents are likely to increase charitable giving.

1. INTRODUCTION

Individuals contribute over $60 billion to charity annually. /1/ While such gifts compete directly with transfers to children, little is known about the effects of the circumstances of the children or the role of bequest taxes on lifetime giving. Given the alternatives to charitable contributions, a properly specified demand function for such contributions should include the price of bequests in addition to the price of the contribution itself [Boskin (1976)]. It should also include the income of children since parents' decisions may be affected by their children's well-being [Becker (1974)].

While intergenerational transfers and their motivation have attracted considerable interest in recent years, they have yet to be incorporated in studies of charitable giving. The current literature effectively assumes that consumption is the only substitute to giving. Altruism, however, is not necessarily limited to charitable institutions. As "charity begins at home," parents may evaluate giving in light of the circumstances of their children and the tax treatment of transfers to them.

This paper evaluates the effects of income, estate and gift taxes, and the income of children on the lifetime charitable contributions of parents. We use a unique data set that matches the federal income tax returns of a sample of wealthy individuals with their estate tax returns and also with the income tax returns of their children. We model the determinants of charitable contributions as a function of the income tax price and the attributes of the donors, including their disposable income, the income of their children, as well as the tax price of bequests.

Section 2 of the paper provides a brief review of the literature on charitable contributions. In Section 3, a model of lifetime giving to charity is presented. Section 4 describes the data sources and the construction of variables. Empirical results are reported in Section 5. These results show that both the income of children and the tax price of bequests affect parents' lifetime charitable contributions. Concluding comments are provided in Section 6.

2. PREVIOUS RESEARCH ON CHARITABLE CONTRIBUTIONS

More than 30 studies have investigated the determinants of lifetime charitable giving. Various researchers have used cross- section samples of tax returns, panel samples of tax returns, surveys, and aggregated time series data to estimate price and income elasticities of giving. Clotfelter (1985) and Steinberg (1990) provide extensive reviews of this literature.

While the studies are too numerous to review in detail, it is worthwhile to note the findings of some of the early studies that set the pattern for later ones. Feldstein and Clotfelter (1976) used the 1963 Federal Reserve Board Survey of Financial Characteristics of Consumers to estimate price and income elasticities. They found that the price elasticity of lifetime contributions ranged from -1.07 to -1.57 and the income elasticity ranged from 0.79 to 0.99. They also found a wealth elasticity of 0.095, identical to that in Steuerle (1987). In a similar study using income tax data for 1970 and 1962, Feldstein and Taylor (1976) found price elasticities of -1.09 and - 1.28 and income elasticities of 0.76 and 0.70.

In general, studies of lifetime contributions have used the log-linear functional form in which the coefficients represent constant price and income elasticities. The price of giving is defined as 1 minus the marginal tax rate where the tax rate is a first dollar tax rate assuming that contributions are initially zero. Most of the studies refine the price variable by adjusting for the possibility of contributing property that has appreciated in value and thereby also saving the capital gains tax. Income is generally measured as disposable income or income less the taxes that would have been paid in the absence of giving. These definitions make the price and income variables independent of the actual amounts of charitable gifts.

Additional variables used varied among the studies, but generally included age, marital status, and family size (or number of dependents). Additional refinements in some of the studies included alternative functional forms [Clotfelter and Steuerle (1981), Lankford and Wyckoff (1990)], government spending variables to measure crowding-out effects [Kingma (1989), Schiff (1985)], volunteer effort [Brown and Lankford (1992), Menchik and Weisbrod (1987)], and tax evasion [Slemrod (1989)].

The empirical studies of charitable giving have generally examined lifetime contributions in isolation from intergenerational transfers. At the theoretical level, however, Boskin (1976) fully accounts for transfer taxes, and posits charitable contributions as a function of estate and gift taxes in addition to the traditional variables used in the literature.

3. THEORETICAL FRAMEWORK

In a simple model of utility maximization, an individual's utility is determined by own consumption (C), the wealth or resources of child heirs (Wh), lifetime charitable contributions (CC), and charitable bequests (CB). The individual's lifetime preferences, with appropriate discounting, can be characterized by the following utility function,

U = U(C, CC, CB, Wh) (1)

where Wh, the heir's wealth or resources, is defined as the sum of own income, gifts (G) and bequests (B) from parent, or:

Wh = Yh + B + G (2)

where Yh is the heir's disposable income.

The individual faces a budget constraint which requires that the sum of consumption (C), charitable contributions and charitable bequests (CC and CB), and gifts (G) and bequests (B) to children not exceed the individual's after-tax income (Y) plus wealth (W), or:

PCC + PCCCC + PCBCB + PBB + PGG < OR = W + Y (3)

where P denotes price. The individual selects the amounts of consumption and transfers to heirs and charity by maximizing (1) subject to (2) and (3). Solving for the first-order conditions, and invoking the implicit-function theorem, the following charitable contribution function can be obtained:

[equation omitted]

Lifetime charitable contributions thus depend on the prices of such contributions, charitable bequests, and gifts and bequests to heirs. In addition, the income and wealth of the donor and the income of the heirs would also affect contributions.

Given income tax rate, [equation omitted], estate tax rate, e, and gift tax rate, g, and normalizing on the price of consumption, the price variables in (4), similar to those in Boskin (1976), are:

[equation omitted]

where [equation omitted] is the expected appreciation rate on the transferred asset and i is the discount rate.

The price of a deductible cash contribution is one minus the marginal income tax rate [equation omitted]. The tax treatment of contributions of appreciated property, however, is more complex because such gifts allow avoidance of the capital gains tax on the accrued gain as well as a deduction for the value of the asset. A more detailed discussion of this treatment is provided in the next section.

Given the estate tax deduction, the price of a charitable bequest depends only on the appreciation rate (n) and the discount rate (i). In order to provide a lifetime gift to a child, the parent will have to pay the gift tax (g) in addition to the gift and therefore the price of a one dollar gift is 1 + g. The current cost of providing a real dollar of inheritance is affected by the estate tax rate (e), asset appreciation rate [equation omitted], and the discount rate (i).

4. DATA AND CONSTRUCTION OF VARIABLES

4.1 DATA

A special study by the Statistics of Income Division of the Internal Revenue Service matched 1981 U.S. federal income tax returns for 1982 decedents to the income tax returns of their children heirs. The decedents income tax records were further matched against their 1982 estate tax records. The combined income and estate tax data set contains information on income sources and deductible expenses reported on income tax records in 1981, as well as the size and composition of terminal wealth reported by their estates in 1982. Demographic information is available on age, marital status, and number of children.

4.2 TAX PRICES

The price of lifetime charitable giving of cash is defined as one minus the marginal tax rate. For taxpayers with significant amounts of corporate stock or real estate other than principal residences, a further adjustment was made to capture the capital gains tax avoided on contributions of appreciated property. The price of contributions of appreciated property is [equation omitted] where [equation omitted] and [equation omitted] are the ordinary and capital gains tax rates, respectively, G is the accrued gain on the asset, V is the market value of the asset and alpha reflects a discounting factor for the expected holding period if the asset were not donated and the possibility that the asset might escape taxation entirely if the asset is held until death. In order to facilitate comparison to previous studies, aG/V was assumed to equal 0.5. /2/ For taxpayers with stock or real estate, the price of giving was a weighted average of the price of giving cash and the price of giving appreciated property with the weights determined by the average proportion of non-cash gifts by income class. Taxpayers were deemed to have significant amounts of appreciable property capable of being contributed if they either received $200 in dividends or rental income (loss) or reported at least $10,000 in corporate stock or real estate on the estate tax return. The minimum price is 0.4, or (1 - 0.50 - 0.50 * 0.20), with an ordinary income tax rate of 0.50 and a capital gains tax rate of 0.20.

In measuring the charitable bequest price, we assume that the transferred asset appreciates at the same rate as the discount rate. The resulting charitable bequest price is one. Estate and gift taxes on intergenerational transfers are combined under the federal unified transfer tax. Under this tax, inter-vivos gifts and testamentary bequests are subject to a unified tax rate schedule. More importantly, the tax liability is based on cumulative transfers. Inter-vivos gifts, for instance, after an annual exemption of $10,000 ($3,000 prior to 1982) per donee, are added to gifts made in prior years in computing the gift tax, with an adjustment (credit) for previously paid taxes. At death, cumulative gifts are added to the estate in computing the estate tax liability and a credit is provided for previously paid gift taxes. The marginal "unified" tax rate is computed by adding $1,000 to the gross estate. The tax price of bequests to children is then defined as the inverse of one minus the computed tax rate. /3/ The maximum price in 1982, the year of death, is 2.86, or 1/(1-0.65), and the minimum is one. Because gift and bequest taxes are part of a unified tax system, it is difficult to estimate their separate effects, and therefore the gift tax price is dropped. /4/

4.3 INCOME AND WEALTH MEASURES

Income of the contributors (decedents) is defined as Adjusted Gross Income (AGI) less the income tax liability computed in the absence of charitable contributions. The income of the children is the mean AGI of non-dependent (i.e., adult) children, also net of income taxes. Observations with incomplete matches for non-dependent children were deleted from the sample in equations that include children's income. For observations where all children are dependents, children's income is assumed to be zero as they have no reported current income. Because the assumption of zero income does not reflect their expected income when they join the labor force, we also provide estimates with a sample that omits observations with dependent children to evaluate the robustness of our estimates.

Wealth is defined as total assets less debts reported on estate tax returns. Several modifications are made to the reported value of total assets. The market value of farms and other small businesses eligible for special estate tax valuations is substituted for the estate tax value, and the share of jointly owned assets excluded from the estate is added back, along with lifetime charitable contributions reported on income tax returns. In addition, life insurance policy proceeds in the estate are excluded. When bequest taxes are considered, wealth is reduced by the amount of the transfer tax to be consistent with the construction of the income measure. This "disposable" wealth represents the maximum amount that can be inherited by the children.

The means for selected variables are reported in Table 1. The first column in Table 1 shows 5,585 observations for primary taxpayers with mean contributions of $14,680, disposable income of $96,280, wealth of $2.7 million, and age of 76 years, with 28 percent over the age of 85. Over 90 percent of the sample report charitable contributions. About 45 percent of the individuals are married, and 43 percent widowed, with the remainder either single (never married) or divorced. The average income tax price is 0.4789. The second column of Table 1 reports mean values for the subsample of parents only. The results are similar to those in the first column. The average individual has about two children, and faces an average tax price of bequests of 1.739. Column 3 presents the means with the sample restricted to observations where the income of adult or independent children is known. Column 4 omits observations where dependent children are present as their current income is zero and their future income is not known. Note that the means for contributions, income, wealth and tax prices in the restricted samples are quite similar to those in the full sample.

5. EMPIRICAL RESULTS

The empirical results on the determinants of lifetime charitable contributions are reported in Table 2. While our main objective is to examine how the charitable giving of parents varies with the income of children and the tax price of bequests, we begin with the full sample in order to see whether our sample of older and wealthier individuals can replicate the results of earlier studies. Tobit estimates for the determinants of charitable contributions are shown for the entire sample in the first two columns, and for parents in columns (3) through (6). The variables are specified in natural logarithms. We add one dollar to the contribution and income variables to avoid taking the log of zero. Column (1) in Table 2 presents an equation similar to that estimated in many previous studies using income tax data. /5/ The results are consistent with the findings in earlier studies which is reassuring given that our sample is more representative of older and wealthier individuals. The income tax price of giving is significant in the equation and implies a price elasticity of -1.27 evaluated at the mean. About 6 percent of this response is due to changes in the likelihood of giving and the rest is due to changes in the amount of giving of those who do give. /6/ The income elasticity of contributions is 0.78. Giving is highest for married taxpayers. Those between the ages of 45 and 85 have higher levels of contributions than those younger and older. The presence of dependent children is not statistically significant.

Column (2) of table 2 adds the wealth variable to the equation in column (1). This reduces the absolute value of the price elasticity from -1.27 to -1.1, with the remaining coefficients essentially unchanged. This suggests that the omission of wealth from charitable giving regressions may lead to overestimates of the responsiveness of contributions to the price of giving. This should not be surprising since non-cash gifts account for approximately 30 percent of the contributions of donors in the sample. /7/ The elasticity of contributions with respect to wealth is about 0.4, significantly larger than that reported in Feldstein and Clotfelter (1976) and Steuerle (1987). One possible explanation of the higher wealth elasticity is that the measure of wealth used in this study is broader than the measures used in the earlier studies.

Focusing on parents only, column (3) of Table 2 replicate the estimates in the second column. The results for parents are generally similar to those obtained for the entire sample. The price elasticity of lifetime contributions is -1.15, the income elasticity is 0.78 and the wealth elasticity is about 0.45. In column (4), the price of charitable bequests is added. The price, income and wealth elasticities are slightly smaller in this equation at -1.12, 0.77 and 0.39, respectively. The coefficient on charitable bequests is statistically significant and positive with an estimate of about 0.6. This implies that through the effect on the price of bequests to children, higher estate tax rates are associated with higher lifetime contributions. Thus, the repeal of bequest taxes, which lowers the price of bequests and increases disposable wealth, would reduce lifetime charitable giving by about 12 percent.

Column (5), which omits observations where the income of adult children is not known, adds the mean income of children and the bequest tax price to column (4) of Table 2. /8/ The estimates in column (5) show an income tax price elasticity of -1.07, income elasticity of 0.71, wealth elasticity of 0.53 and bequest price elasticity of 0.75. The elasticity of charitable giving with respect to the income of the children is positive and statistically significant, with an estimated coefficient of 0.08, indicating that more will be contributed to charity when the children are more well- off. /9/ These results imply that giving increases by one percent when either the children's income increases by 12 percent or the donor's income increases by 1.5 percent.

Column (6) replicates the estimates in column (5), but omits observations where dependent children are present. For this smaller, and somewhat older group the estimated income tax price elasticity is -1.16. The income elasticity of 0.58 is somewhat lower than in the other equations and the wealth elasticity of 0.63 is somewhat higher. The effect of children's income on lifetime giving is larger with an elasticity of 0.14, /10/ while that of the bequest price is lower with an elasticity of 0.65. /11/ The larger effect could be the result of the difference in the samples or the inability to properly measure the income of dependents in the column (5) specification. For this sample, charitable gifts are likely to increase by one percent when the children's income increases by 7 percent or the parent's income increases by 1.7 percent.

5.1 CHARITABLE BEQUESTS

So far the paper has focused on the effects of children's income and bequest taxes on lifetime charitable giving. A case can be made for also evaluating the impact of the effects of these variables on charitable bequests at death. After all, if children's income influences the lifetime giving of parents, one might expect a similar effect on charitable bequests. However, a full analysis of charitable bequests and the role of children's income would require the estate tax returns of both parents, and their income history over a number of years along with initial wealth endowments (inheritances). Unfortunately, our data on charitable bequests contains information only for one parent, unlike lifetime contributions which are generally reported jointly by both parents. We also do not observe the income or tax price of the surviving spouse or how that spouse will disburse the remainder of the estate. Nevertheless, with the caveat on data limitations in mind, we present estimates for charitable bequests in Table 2. The specification is similar to those previously used in studies of charitable bequests except that we add the income of the children. /12/

The mean charitable bequest for parents with non-dependent children is $346,194, with about 25 percent of the observations reporting charitable bequests. /13/ This compares to mean lifetime contributions of $18,546 with over 90 percent reporting such contributions. Column 7 of Table 2 shows the estimation results for charitable bequests. This model differs from the model for lifetime contributions in several respects. Relative to the lifetime definition, wealth is adjusted to account for resources available at death by excluding the surviving spouse's share of joint property, adding back life insurance proceeds, and excluding lifetime charitable contributions. /14/ The lifetime income and price variables are excluded from this model. /15/ The elasticity of bequests with respect to wealth is about 1.3, implying that charitable bequests increase more than proportionally with the size of the estate. The bequest tax price has a strong effect on charitable bequests with an elasticity of -2.5 evaluated at mean values. This result is within the range of estimates of previous studies. /16/ About 24 percent of the response is due to the change in probability of giving. While children's income is positively related to charitable bequests, the coefficient is not statistically significant.

To shed light on the effect of children's income on charitable bequests, we further split the sample into married and widowed decedents. For married decedents, the Tobit equation (not reported) show an estimated coefficient for children's income of -0.49 with standard error of 0.666. For widowed decedents, the estimated coefficient is 0.173 with standard error of 0.332. Theses coefficients suggest that children's income has a small influence on the charitable bequests of parents, as both estimates are imprecisely measured. Again, this may be attributable to a lack of longitudinal data. It is also possible that testamentary and inter-vivos transfers are governed by different factors.

6. CONCLUSIONS

This paper extends the analysis of charitable contributions to take into account intergenerational transfers. Previous studies of charitable giving have generally ignored the effects of bequest taxes and the income of the heirs. Using income and estate tax records of a sample of affluent individuals, the paper investigates the effects of the tax price of bequests, wealth, and the income of children on charitable giving. One caveat is that since the sample reflects older and wealthier households, the results may not apply to the whole population. However, the estimated income and tax price elasticities are similar to those in previous studies of the whole population.

The results suggest that lifetime charitable giving and intergenerational transfers are substitutes. Transfer taxes are found to affect lifetime contributions with an estimated elasticity of contributions with respect to the tax price of bequests of about 0.6. The paper also finds that the income of children affects the amount that parents contribute to charity. The results show positive elasticity coefficients of up to 0.14 for contributions by parents with respect to the mean after-tax income of the children, compared to an elasticity of 0.6 or 0.7 with respect to parent's own income. The results imply that lifetime contributions would increase by one percent if children's income increases by 7 to 12 percent or parent's income increases by 1.5 to 1.7 percent.

While the results show that intergenerational transfers are an important consideration in lifetime charitable giving, their exclusion does not appear to have biased the findings on the income tax price of contributions in previous studies. However, the omission of wealth may have produced an upward bias in previous estimates of tax price and income elasticities.

Becker, G.S. 1974, A Theory of Social Interactions, Journal of Political Economy 82, 1063-1093.

Boskin, M.J. 1976, Estate Taxation and Charitable Bequests, Journal of Public Economics 5, 27-56.

Brown, E. and H. Lankford. 1992, Gifts of Money and Gifts of Time: Estimating the Effects of Tax Prices and Available Time, Journal of Public Economics 47, 321-341.

Clotfelter, C.T. 1985, Federal Tax Policy and Charitable Giving (University of Chicago Press, Chicago).

Clotfelter, C.T. and C.E. Steuerle. 1981, Charitable Contributions, in: H. Aaron and J. Pechman, eds., How Taxes Affect Economic Behavior (Brookings Institution, Washington, D.C.) 403-446.

Feldstein, M. 1975, The Income Tax and Charitable Contributions: Part I -- Aggregate and Distributional Effects, National Tax Journal 28, 81-100.

Feldstein, M. and C. Clotfelter. 1976, Tax Incentives and Charitable Contributions in the United States: A Microeconometric Analysis, Journal of Public Economics 5, 1-26.

Feldstein, M. and A. Taylor. 1976, The Income Tax and Charitable Contributions, Econometrica 44, 1201-1221.

Internal Revenue Service, 1993, SOI Bulletin 13, No. 2, (Washington, D.C.).

Internal Revenue Service, 1983, Statistics of Income - 1981, Individual Income Tax Returns (U.S. Government Printing Office, Washington, D.C.).

Joulfaian, D. 1991, Charitable Bequests and Estate Taxes, National Tax Journal 44, 169-180.

Kingma, B. 1989, An Accurate Measurement of the Crowd-out Effect, and Price Effect for Charitable Contributions, Journal of Political Economy 97, 1197-1207.

Lankford, R.H. and J.H. Wyckoff. 1990, Modeling Charitable Giving Using a Box-Cox Standard Tobit Model, Review of Economics and Statistics, 460-470.

Menchik, P. and B. Weisbrod. 1987, Volunteer Labor Supply, Journal of Public Economics 32, 159-183.

Schiff, J. 1985, Does Government Spending Crowd Out Charitable Contributions?, National Tax Journal 38, 535-546.

Slemrod, J. 1989, Are Estimated Tax Elasticities Really Just Tax Evasion Elasticities? The Case of Charitable Contributions, Review of Economics and Statistics 71, 517-522.

Steinberg, R. 1990, Taxes and Giving -- New Findings, in The Nonprofit Sector in the United States and Abroad, Proceedings of the 1990 Spring Research Forum.

Steuerle, E. 1987, Charitable Giving Patterns of the Wealthy, in T. Odendahl, ed., America's Wealthy and the Future of Foundations (The Foundation Center, New York).

TABLE 1: MEANS OF SELECTED VARIABLES

Entire Parents Sample Item Sample All Sample 1 Sample 2

______________________________________________________________________

Contributions 14,680 16,775 17,181 18,546 In Contributions 6.905 6.987 7.007 7.039

Disposable AGI 96,280 94,599 95,071 96,533 In Disposable AGI 10.970 10.969 10.982 11.021

Wealth 2,731,653 2,757,206 2,841,275 2,878,642 Disposable Wealth -- 2,041,748 2,103,915 2,081,614 In Wealth 14.405 14.410 14.410 14.420 In Disposable Wealth -- 14.156 14.150 14.148

Children -- 2.250 2.129 2.146 Number of dependent children -- 0.171 0.195 0.096 Number of others (independent) -- 2.079 1.934 2.050 Dependents present (yes= 1) -- 0.103 0.125 0.000 Mean Disposable AGI -- -- 50,865 57,082 In Mean Disposable AGI -- -- 10.133 10.476

Age 75.99 76.46 76.170 77.444 Age under 45 0.018 0.011 0.013 0.000 Age 45 to 54 0.050 0.045 0.053 0.002 Age 55 to 64 0.116 0.108 0.108 0.032 Age 65 to 74 0.212 0.217 0.206 0.103 Age 75 to 84 0.320 0.336 0.340 0.352 Age 85 or over 0.248 0.283 0.280 0.320

Marital status Married 0.454 0.412 0.406 0.389 Widowed 0.427 0.524 0.526 0.549 Single 0.071 0.004 0.006 0.005 Divorced/Separated 0.047 0.059 0.062 0.056

Tax Prices (x100) Lifetime contribution 74.89 47.72 47.89 46.80 Bequest -- 173.90 174.99 177.61

In Tax Prices Lifetime contribution 3.795 3.791 3.794 3.773 Bequest -- 5.131 5.139 5.155

Observations 5,585 2,868 1,979 1,732 Observations with Contributions 5,023 2,607 1,803 1,582 ______________________________________________________________________TABLE 2: TOBIT ESTIMATES FOR CHARITABLE GIFTS

Lifetime Contributions Charitable

Entire Sample Parents Only Bequests Explanatory (1) (2) (3) (4) (5) (6) (7) Variable ______________________________________________________________________ Constant 2.414 -3.285 -4.294 -6.565 -9.515 -9.561 -14.329 (0.747) (1.137) (1.478) (1.817) (2.202) (2.458) (20.650)

In Income Tax Price -1.312 -1.174 -1.216 -1.162 -1.108 -1.212 -- (0.124) (0.125) (0.164) (0.166) (0.197) (0.216) --

In Income 0.780 0.748 0.783 0.773 0.709 0.590 -- (0.036) (0.034) (0.047) (0.047) (0.053) (0.062) --

In Net Worth -- 0.396 0.467 0.396 0.535 0.641 3.945 -- (0.060) (0.077) (0.084) (0.109) (0.126) (0.826)

Married 0.933 0.759 0.541 0.682 0.670 0.696 -1.188 (0.086) (0.089) (0.116) (0.133) (0.160) (0.177) (1.063)

Age under -0.399 -0.284 -0.161 -0.146 0.267 -- -- 45 (0.318) (0.316) (0.546) (0.546) (0.649) -- --

Age 45 to 0.379 0.401 0.125 0.179 0.274 -0.269 -9.975 54 (0.207) (0.205) (0.310) (0.310) (0.368) (0.635) (4.609)

Age 55 to 0.474 0.482 0.507 0.560 0.447 0.388 -9.876 64 (0.147) (0.146) (0.199) (0.200) (0.241) (0.272) (1.930)

Age 65 to 0.514 0.497 0.469 0.500 0.457 0.418 -6.823 74 (0.117) (0.116) (0.151) (0.151) (0.183) (0.193) (1.193)

Age 75 to 0.519 0.518 0.504 0.514 0.331 0.309 -3.986 84 (0.102) (0.101) (0.132) (0.132) (0.158) (0.163) (0.927)

Dependent Children present 0.065 0.065 0.078 0.082 0.250 -- -- (yes= 1) (0.190) (0.188) (0.211) (0.211) (0.246) -- --

In Child -- -- -- -- 0.081 0.149 0.015 Income -- -- -- -- (0.037) (0.053) (0.301)

In Bequest -- -- -- 0.606 0.750 0.667 -11.263 Tax Price -- -- -- (0.282) (0.351) (0.399) (2.752)

o 2.891 2.874 2.714 2.711 2.693 2.752 12.582 (0.030) (0.030) (0.039) (0.039) (0.046) (0.050) (0.518)

Likelihood -13177 -13155 -6638 -6635 -4576 -4054 -2,246

Observations 5,585 5,585 2,868 2,868 1,979 1,732 1,732 _____________________________________________________________________Standard errors are reported in parentheses.

FOOTNOTES

/*/ Published in the Journal of Public Economics (1996) 55-68.

/**/ We thank Charles Clotfelter, Hamilton Lankford, Harvey Rosen, Gabriel Rudney, Eugene Steuerle, and two anonymous referees for useful suggestions. The views expressed are those of the authors and do not necessarily reflect those of the Department of the Treasury.

/1/ See Internal Revenue Service (1993, p. 145). This amount is for individuals with itemized deductions only.

/2/ This value was first used by Feldstein and Clotfelter (1976) based on a maximum likelihood search procedure. Later studies of charitable contributions have generally used this value.

/3/ Capital gains taxes may affect the price of bequests to children if the asset held during life contained appreciated value. This is because these taxes would be paid if the parent were to sell the property to finance consumption. In this case the price of bequests becomes [equation omitted], where tc is the capital gains tax rate and b is the ratio of accrued capital gains to market value. The implications of this are addressed in the empirical section. Note that capital gains taxes may also affect the price of charitable bequests under this scenario.

/4/ Even if one were able to disentangle the gift tax rate from the estate tax rate, the tax prices are likely to be highly correlated even though the capital gains tax treatment of transferred property differs. Note, however, that the non-tax factors, such as control of a family business, can affect the implicit "prices" of gifts versus bequests.

/5/ The logarithmic specification is supported by the Box-Cox estimates in Lankford and Wykoff (1990), who find that the preferred model is very close to the logarithmic model.

/6/ See McDonald and Moffitt (1980) for discussion of this decomposition.

/7/ In 1981, non-cash gifts accounted for 15 percent of the $29.8 billion in total charitable contributions, but over half of the contributions of taxpayers with at least $200,000 of Adjusted Gross Income [Internal Revenue Service (1983), p.54].

/8/ We were not able to obtain income tax records for these heirs. Recall that income of children is assumed to be zero when all children are claimed as dependents on the tax return of the parent.

/9/ We obtained the same 0.08 coefficient when we added child income to the equation estimated for the larger sample in column 4 and used dummy variables for non-matches.

/10/ In an alternative specification that is not reported, children's income is measured by adjusting their income to age 45 assuming a 4 percent annual growth rate of income. The estimated coefficient for children's income was 0.161 with a standard error of 0.035.

/11/ The coefficient on the price of bequests rises to 0.9 when capital gains taxes are combined with the estate tax.

/12/ We expanded on these specifications by including income and income tax price of lifetime giving. However, the estimated coefficients on these variables were statistically insignificant. Ideally, we would have preferred to replace the wealth variable with a measure of total lifetime income to examine how parents allocate charitable gifts between lifetime giving and bequests.

/13/ Since married decedents generally leave much of their wealth to their surviving spouse, the reported fraction of returns with charitable bequests may understate the relative frequency of giving by households.

/14/ The results are invariant to the inclusion or exclusion of lifetime charitable contributions in the definition of wealth.

/15/ We experimented with adding income and tax price of lifetime contributions to this equation. While the coefficients for both exhibited the theoretically correct sign, neither was statistically significant. Of course, this should not be surprising since both are measures for both parents, while charitable bequests are known only for the decedent parent.

/16/ For 1957-59 and 1969 estate tax returns, Boskin (1976) found charitable bequest tax price elasticities of -1.2 and -2.0, respectively. Using 1976 estate tax returns, Clotfelter (1985) found tax price elasticities of -1.67 and -2.79 depending on the definition of wealth. Joulfaian (1991) found a tax price elasticity of -3.0 for charitable bequests.

END OF FOOTNOTES

$include:RR_NAL-talink$

## Add comment

## Comments