As a follow-up to a recent presentation of my financial disaster course, I have conducted further investigations as to where the peak of the Laffer Curve might actually reside.  As discussed in the presentation, the curve is typically drawn with the peak at 50% as shown in Figure 1.

Figure 1: Typical depiction of Laffer Curve (rotated 90 degrees)

What is often not discussed is that this curve implies that the overall economy shrinks with the tax rate. The total economy for a given tax rate can, in fact, be determined mathematically if the shape of the Laffer Curve is known. For example, the typical Laffer Curve would produce an economic image as shown in Figure 2; the height of the lighter color represents the money retained by the people and the darker color represents the amount of money taken by the government.

Figure 2: Implied Economy

Note that some people call the peak of the Laffer Curve the "optimum," this is not really an accurate term; the optimum would be the location with maximal societal revenue while maintaining essential government services (i.e., protection of the system). This would certainly be to the left of the peak. The peak simply represents where government revenues are highest.

It should be noted that the Laffer Curve only precisely defines two points: one at 0% taxation and 0 revenue and the other at 100% taxation and 0 revenue. It also postulates that any point in between would result in positive revenue; however, the actual shape of the curve is not known. This has led to some speculation as to where the peak of the curve is.

In reality, this is a very complex question as the point at which it peaks probably is different for different types of taxes (e.g., capital gains taxes versus income taxes) and there are many inter-relationships (e.g., the tax rate represented in the curve is a cumulative tax rate including federal, state, local, payroll, etc.) Nonetheless, we can look to empirical evidence to try to determine where we might be at present for the federal tax rate. The OMB publishes a table that records the revenues received from the income tax as a percentage of GDP. This can be matched against historical tables of the top tax rate in order to produce the following graph showing data points for all years since World War II.

Figure 3: Government Revenue vs. Tax Rates

A rough statistical analysis on this chart indicates that there is considerable confidence that tax rates of 70% and above result in lower net revenue, thereby agreeing with the concept of the Laffer Curve. NOTE: The figure only considers the top tax rate on the X axis but considers all income tax revenue on the right axis. It is almost certain that if we only assumed income from the highest earners that we would see the data points on the right drop even more substantially. 

The chart also appears to suggest that the peak revenue occurs at a top federal tax rate of roughly 40%, although a statistical analysis of the data reveals that the confidence level of such a statement is somewhat weak, it is still the best guess from the chart; to pinpoint this further, we will need to wait for additional data points.

If one combines the 40% federal tax rate with other federal taxes and credit phase outs and with state and local taxes, one realizes that this 40% federal tax rate likely corresponds to a 60-70% effective marginal tax rate (i.e., if a person earns an extra dollar, they end up paying $0.40 in federal income taxes plus another $0.20 - $0.30 in other federal, state, and local taxes). This corresponds with what many economists have predicted as the peak of the curve (of course, they have a range and many on the left will try to predict 70% as the top federal income tax rate - but the real-world data from above simply do not support that claim).

So what does that mean in practical terms? If the Bush tax cuts were to expire, it would not be the end of the world; it would most likely increase federal revenues, but it would also significantly hurt the economy - and as a result, it would also decrease state and local revenues. However, I also looked at the data and realized that the amount of additional revenue is likely half of what is often cited. At best, we could expect revenues to increase by $200 billion per year versus the widely publicized predictions of $300 - $400 billion per year.  Further, the benefits of these cuts are more than likely heavily concentrated on the extreme lower end (i.e., changes to the child tax credit, etc) and only eliminating the cuts on the highest earners would likely result in less than $40 billion per year rather than the $70 billion per year claimed. 

In short, we are probably near the sweet-spot for tax revenues given our current tax code. There are likely ways to increase revenues (e.g., shifting to a consumption tax or a flat tax), but as long as we use the current tax code design, we probably will never sustain federal income tax revenues of more than 10% of GDP.