Doing Well by Doing Good

Bertrand, Marianne, Matilde Bombardini, Raymond Fisman, and Francesco Trebbi. "Tax-Exempt Lobbying: Corporate Philanthropy as a Tool for Political Influence." American Economic Review, forthcoming.

It seems safe to say that corporate philanthropy, in general, is beneficial to society. It appears to be a "win-win" situation where corporations engage in prosocial behavior and in return receive good publicity that may increase longer-term profits. But if some corporate giving is in part a means of gaining political influence with elected officials, then taxpayers are, in effect, subsidizing these tax-deductible contributions — and voters and investors are losing the transparency afforded by regulations on political contributions.

A forthcoming article in the American Economic Review by Marianne Bertrand of the University of Chicago, Matilde Bombardini and Francesco Trebbi of the University of British Columbia, and Raymond Fisman of Boston University suggests that some corporate philanthropy is politically motivated and attempts to quantify the amount.

The authors approach their research question with three different strategies. They first use data on Fortune 500 and S&P 500 companies' charitable contributions and PAC contributions to establish that for a given corporation and congressional district, there is a positive relationship between the firm's charitable contributions and political action committee (PAC) spending. They show that the movement of charitable contributions over time looks very similar to that of contributions to PACs, a more traditional channel of political influence.

The authors' second strategy is to use that data to show that a firm's charitable contributions are more likely to go to the congressional districts of representatives who serve on committees that are of interest to the firm. First, they use lobbying reports to assemble a variable that records the number of issues covered by the congressional committees on which a representative serves that are of interest to a corporate foundation during a certain congressional session. They then use regression analysis to test their theory. If their hypothesis is correct, then a higher number of issues of interest covered by a representative from a certain district should be associated with a larger contribution to that district's charities from the company's foundation.

Yet establishing a positive relationship between issues of interest and charitable contributions is not sufficient to prove that issues of interest cause charitable contributions to increase. It could be that corporations donate for nonpolitical reasons in locales whose representatives, due to the issues important to their constituency, self-select into committees of interest. Establishing causation by eliminating alternative explanations is what economists call "causal identification." That is why the authors use "fixed effects" regressions, a type of regression that takes into account such hidden factors. By using this method, the authors are able to capture the increases and decreases in charitable contributions associated with a representative joining or leaving a committee of interest. They argue that since it is unlikely that any other alternative would explain why the amount of contributions change over time, the effect they capture is likely to be causal. Using a few different versions of this empirical model, they find that a 1 percent increase in issues of interest covered by a district's representative leads to a 0.04 to 0.091 percent increase in charitable contributions to that district.

The third strategy they use is to link data on representatives' personal connections to nonprofits with data on charities to which firms' foundations donate. (Representatives' personal connections to nonprofits must be disclosed under the Ethics in Government Act of 1978.) They use these data to explore whether firms are more likely to donate to a nonprofit with a connection to Congress if it aligns with their lobbying interests. To do this, they create measures of a nonprofit's political relevance to a firm, including a variable that indicates whether a nonprofit is connected to a politician who serves on a congressional committee of interest to a firm's lobbying efforts. In general, they find that, among nonprofits with a connection to a representative, an increase in a nonprofit's political relevance to a firm increases the likelihood that that nonprofit receives a charitable grant from the firm.

The authors build an economic model to quantify the fraction of corporate philanthropy that is politically motivated. In the model, a congressional committee assignment that is relevant to a firm increases the productivity of investment in politically motivated charitable giving and in PAC contributions. They estimate the share of corporate giving that is politically motivated to be 6.3 percent at its most conservative, which amounts to $1.13 billion of the $18 billion that U.S. firms gave in 2014, the last year of the sample.

This research sheds light on the not-insignificant role that corporate philanthropy plays in the political arena and suggests that corporate giving may not always be entirely a "win-win" situation.