Are Americans Becoming Less Honest? Evidence from Tax Payments

Until recently, I did the adjunct professor thing at an MBA program at a business school in the Los Angeles area – I taught the first year sequence in economics, and an elective advanced statistics class. Once a quarter, I was expected to talk to my class about ethics. Its something I always had a hard time doing – I’ve always been under the impression (if someone else has said this before, forgive me, if not, I’ll be happy to claim this as “cactus’ 1st law”) that those who talk a lot about ethics, morality, or patriotism generally have none.

Maybe its my impression, maybe I’m getting older, but for whatever reason, I’ve gotten the impression that people in this country have been doing a lot more talking about those three topics over the past few decades. This post is an attempt to measure whether that impression is true, in the context of taxes.“> My last post covered tax enforcement. The post showed that, for the period from 1988 to 2003, the greater the share of the budget going to the Treasury Department, the greater the share of the total individual income taxes paid by the top 1% and 5% of income earners. In other words, as enforcement increases, those with more discretion in calculating the taxes they owe end up paying more of it.

A more direct measure of the government’s willingness to enforce tax laws is the“> IRS’ operating budget as a share of the total federal budget“> (NIPA Table 1.1.5) .

This can be compared to the amount of“> corporate, individual, employment, estate, gift, and excise taxes as a percentage of GDP. Presumably, if Americans are becoming less honest with respect to their taxes, over time we would expect to see, over time, a greater degree of correlation between enforcement and amount paid, especially in those types of taxes of taxes in which cheating is easiest.

Without running any numbers, I would guess taxes, ranked in order of ease of cheating, look like:

1. Corporate – consider transfer pricing and any number of other gimmicks
2. Individual – easy to cheat at for high income folks, not so easy for wage earners
3. Excise taxes – phantom sales
4. Estate taxes – easy to cheat on, but due to other requirements (namely, a death), may be lumpy and thus slightly less related to time
5. Gift taxes – easy to cheat on, but why not simply place give the giftee a “consulting” fee?
6. Employment taxes – hardest to cheat, since it requires the connivance (or at least the bamboozling) of the employees as well as the government.

The table below shows the correlation over four decades between the IRS operating budget and corporate, individual, and excise taxes as a percentage of GDP.


Clearly, over time, the correlation between corporate taxes and the IRS budget is increasing quickly over time. The greater the IRS budget, the more corporations pay in taxes, in general. The same seems to be true of individual and excise taxes, albeit each of these has one hiccup along the way in the 1980s. (As per my earlier post on taxes, focusing on the relationship between high income individuals’ taxes and the IRS budget would probably show stronger results.)

The next table shows


As with individual and excise taxes, estate taxes do seem to be getting more correlated with the IRS budget, with the exception of the 1980s. Gift taxes also seem to be showing more correlation over time, but the relationship is weak. As to employment taxes, it is very hard to reach much of a conclusion.

Thus, it seems that many taxpayers, corporate and individual, are deciding how much to pay in taxes in large part based on much funding the IRS receives. Put another way, over time Americans have become less likely to be honest with their taxes without threat of punishment.

And if you’re wondering…. the IRS’ share of the budget hit a maximum of 1.48% in 1999. In 2005, it was down to 1.14%.