If you think you wound up paying too much in taxes this year, maybe you ought to hire a lobbyist. Or two. Or 20. After all, it’s a strategy that seems to be working well for some of the nation’s biggest corporations.

As Americans prepare for tax day 2012, a new Sunlight analysis of lobbying and corporate tax rates finds that among 200 of the largest U.S. companies, the companies that spent the most on lobbying most effectively reduced their reported tax rates between 2007 and 2010.

On average, companies we examined reported paying a slightly lower overall tax rate in 2010 than in 2007 (average tax rate of 29.3 percent in 2010 as compared to 29.9 percent in 2007), with a decline in the median reported tax rate from 31.8 percent to 31.6 percent. Fifty-five percent of the companies paid a lower rate in 2010 than in 2007.

But of the eight companies that spent the most on federal lobbying between 2007 and 2009, seven decreased their overall tax rate between 2007 and 2010. And six of the Big Eight enjoyed a decrease of at least seven percentage points.

Table 1. Changes in reported tax rates

Combined, the Big Eight spent $540 million on lobbying between 2007 and 2009. In total, they filed 332 lobbying reports that mentioned taxes, naming 491 different tax bills over the three-year period we investigated (see table 2).

Table 2. Lobbying on tax bills

Compared to what their taxes would have been if their 2007 tax rates were applied to their 2010 income, we estimate that the seven companies that lowered their taxes saved a combined $11 billion on $120 billion in reported 2010 profits. If we assume that the entire reduction was due to their lobbying, the return on investment would be 2,069%.  Of course, this is probably not the case. Without a detailed analysis of these companies’ taxes, it would be impossible to tell why their rates fell. But we can observe that it is very unlikely that the eight companies that lobbied the most between 2007 and 2009 all would have seen such significant drops in their tax rates by random chance alone.

Statistically, the likelihood of that seven of eight firms that ran up the biggest lobbying tabs all lowering their reported tax rates by chance alone is about seven percent, which makes it highly unlikely but within the real of random possibility (assuming we take the overall probability of an individual company lowering its taxes at 55 percent). Moreover, only 19 of the nation’s 200 highest earning companies reduced their tax rate by more than seven percentage points. Within this universe of companies, the likelihood of six of the Big Eight lowering their rates by at least seven percentage points purely by random chance is less than 1 in 100,000.

Of course, the Big Eight were not unique in lobbying on taxes. More than two-thirds of the nation’s 200 largest companies lobbied on at least one tax bill in the last three years. Of the 182 large companies that reported at least some lobbying during the period, 73 percent lobbied on taxes. Combined, the 182 companies filed 2,405 unique reports mentioning tax lobbying, with a total of 3,600 unique tax bill mentions.

Companies lobby so much because Congress is constantly making changes to the tax code. Between 1987 and 2011, the number of pages in the CCH Standard Federal Tax Reporter (one measure of the growing complexity of the federal tax code) more than doubled, growing from 33,030 pages to 72,574 pages. More than half of those pages have been added since 2001.  In 2005, the President’s Advisory Panel on Tax Reform counted approximately 15,000 separate changes to the tax code since 1986 (more than two a day). “Each one of these changes had a sponsor,” wrote the panel in its report, “and each had a rationale to defend it. Each one was passed by Congress and signed into law.”

It’s not surprising then, that we recently found that a committee seat on the House Ways and Means Committee (which has jurisdiction over tax rates) is worth an estimated $258,000 in extra fundraising. After all, a substantial number of companies appear to be after highly-prized tax benefits. And PAC and employee contributions to committee members are generally an effective way of gaining access. 

Regression analysis

Taking the 200 companies as a whole, we estimate that for each additional $1 million that companies spent lobbying between 2007 and 2009, their 2010 tax rate fell by 7/100ths of a percent. While that might not sound like much, for a $2.5 billion company at the median in our sample of large companies, each $1 million spent on lobbying translates into an estimated annual tax savings of $1.8 million – almost double the original investment, with benefits likely to continue in the future. Table 3 reports the results of a OLS regression estimating companies’ 2010 tax rate as a function of their lobbying totals between 2007 and 2009, controlling for size, industry, and 2007 tax rate. The estimate of a decrease of 7/100ths of a percent is statistically significant, though just slightly.

Table 3. OLS regression results of effect of lobbying on tax rate

(coefficients for industry at 2-digit NAICS code not reported) Adjusted r-squared: 0.4697

However, the result appears to be largely driven by the eight companies that lobby the most. If we remove those companies from the sample, the estimated effect of lobbying on tax reduction drops to 5/1000ths of a percent per $1 million spent on lobbying, with a 35.5% chance of being zero (the absolute minimum for statistical significance is a 10% chance of being zero).

Conclusion

When it comes to paying less in taxes, having an army of lobbyists appears to be helpful. Many companies lobby on taxes, but those who spend the most report the largest and most consistent declines in tax rates. Of the eight companies who spent the most money lobbying between 2007 and 2009 seven saw their 2010 tax rates decline from what they paid in 2007. While it is difficult to show causality, the likelihood of this happening by random chance is less than 1 in 100. And of those eight companies, six reduced their tax rates by at least seven percentage points. Given the larger patterns we’ve observed, the likelihood of this happening by random chance is less than 1 in 100,000. At the very least, we know that the companies that lobby the most are also the  companies who have figured out some way to pay millions less in taxes than they did just a few years ago.

Methodology and sample

Our set of 200 companies includes the largest 200 U.S. companies (ranked by 2010 pre-tax income) that met the following criteria: in both 2007 and 2010 they reported tax rates that ranged between -50% and 50% both years, and their income in 2007 was positive. We did this because we wanted to eliminate cases in which tax rates were likely to be driven by one-time events. This is admittedly not a perfect approach, but we wanted tax rates to be appropriately comparable between the two years and to eliminate outliers.

We calculated the base tax rate using Compustat data, and thus we rely on what companies reported in their financial statements as their pre-tax income and their income tax paid (in this, we follow the approach of Richter et al. 2009, who also find that companies that lobby more pay less in taxes). Our decision to look at the cumulative lobbying over three years is based on the assumption that tax lobbying takes place over multiple years and tax benefits do not always kick in immediately.

The list of companies and their tax and lobbying data can be found here.

Special thanks to Alison Rowland for her help in preparing this analysis.

Correction: Altria's tax rates were originally reported incorrectly. The post has been updated to reflect Altria's correct tax rates.