Month: June 2018

In my post, The Education Tax Reduces Inequality and the Incentive to Work, I illustrated how the high cost of college combined with income based pricing have turned education pricing into a tax with potentially significant effects on work and savings incentives. David Henderson pointed me to a paper by Martin Feldstein in 1992, College Scholarship Rules and Private Saving which states the issues very well.

This paper examines the effect of existing college scholarship rules on the incentive to save. The analysis shows that families that are eligible for college scholarships face “education tax rates” on capital income of between 22 percent and 47 percent in addition to regular slate and federal income taxes. The scholarship rules also impose an annual tax on previously accumulated assets. Through the combination of the implied tax on capital income and the associated tax on previously accumulated assets, the scholarship rules that apply to a middle-income family reduce the value of an extra dollar of accumulated assets by 30 cents in four years. A similar family with two children who attend college in succession will see an initial dollar of assets reduced to 50 cents. Such capital levies of 30 to 50 percent are a strong incentive not to save for college expenses but to rely instead on financial assistance and even on regular market borrowing, Moreover, since any funds saved for retirement are also subject to these education capital levies. the scholarship rules discourage retirement saving as well as saving for education. The empirical analysis developed here, based on the 1986 Survey of Consumer Finances, implies that these incentives do have a powerful effect on the actual accumulation of financial assets. More specifically, the estimated parameter values imply that the scholarship rules induce a typical household with a head aged 45 years old, with two precollege children, and with income of $40,000 a year to reduce accumulated financial assets by $23,124 approximately 50 percent of what would have been accumulated without the adverse effect of the scholarship rules.

Dick and Edlin made a similar point in 1997 but, as far as I can tell, there have been only a handful of papers on this issue since that time. The cost of college, however, is about twice as high today as in the 1980s and price discrimination is much more extensive so the effects are likely larger today.

At the NYTimes David Leonhardt breaks families down into six income classes from the poor to the very affluent, defined as follows:

Using a tuition calculator he then estimates tuition (including room and board) by income class at 32 colleges and universities (see below–the darker dots indicate the richer income classes). At many private colleges and universities it is not unusual for some students to be paying $70,000 per year while others pay less than $5,000, for exactly the same education. (Tyler and I provide some similar data on college price discrimination in Modern Principles.) Leonhardt’s point is that a poor student can afford an expensive education and might actually save money by going to an elite private university rather than to a state college.

Price discrimination in education has two other, less recognized, consequences. It reduces income inequality and it reduces the incentive to work. If every firm charged twice as much to someone who earned twice as much, there would no consumption inequality despite high measured income inequality. The rich don’t pay more than the poor when they buy the same basket of goods at the grocery store but they do pay much more for the same education.The Affluent pay approximately $70 thousand for education at the colleges Leonhardt examines while the Upper-Middle pay about half that. The effect on inequality is significant for families with kids in college. An Affluent person is 52% richer than an Upper-Middle income person (186/122=1.52) but an Affluent person with a kid in college is only 33% richer than an Upper-Middle income person with a kid in college (((186-70)/(122-35)=1.33). Shockingly, an Affluent person with two kids in college is actually poorer than an Upper-Middle income person with two kids in college! ((186-140)/(122-70)=0.88.

Under income-based pricing the education tax becomes an income tax with all the negative aspects of income taxes on behavior such as diminished work incentives. Let’s take a closer look at an Upper-Middle income parent earning $122 thousand per year. If this parent gets a promotion or takes on extra work that bumps their salary by $64 thousand, they move from being Upper-Middle income to Affluent. At least on paper. At an income of $122 thousand the parent will be paying approximately $35 thousand to send their child to college but at $186 thousand they will have to pay $70 thousand for the same college so the increase in salary of $64 thousand is an effective increase of only $29 thousand. If the Upper-Middle income parent has two children in college, earning more money actually results in a net loss. For an Upper-Middle income family with two kids spaced in age a few years apart the education tax could be a very severe work disincentive for up to a decade.

The education tax is a peculiar tax as it is often paid to private organizations rather than to the government but it is still a tax and for those of Upper-Middle income the education tax is a very significant tax.

The consequences of the education tax on inequality and work incentives are neither well studied nor well understood.