The Problem with Progressivity

There's an argument in the comments section of this post at Alas about the effects of marginal tax rates on labor supply.

I do think that there's something to the idea that the short-term labor supply is somewhat inelastic. Ronald Reagan's stories of making a few movies and then taking the rest of the year off were plausible in the days of 90% marginal tax rates, but it's much harder to imagine a surgeon or CEO quitting or cutting back on hours if his marginal tax rate goes from 40% to 50%.

But the long-term effects of a steeply progressive income tax may be a different story altogether, because then we have to take into account the effect of the tax structure on future career choices.

For example, to become a doctor takes a lot of work. Four years of college, four of medical school, and then another three to ten of on-the-job training, depending on the specialty. After that, they start making a lot of money. They have to, to make up for the fact that they've missed out on 7-14 years of post-college earning potential, and for the 100-hour workweeks residents have to put up with.

I did the math, and it turns out that my decision to drop out of college at 19 to work as a software engineer actually compares favorably to most types of medicine (except for the most lucrative specialties) in terms of lifetime income potential, even assuming that I never hit it big with stock options or by starting my own company.

Insofar as that's just a result of market forces--if it's because consumers want 45 years of software engineering starting now more than they want 35 years of medicine starting ten years from now--that's fine. But the incentives are distorted by the progressive tax system. Even if a doctor and I made exactly the same amont of money over our lifetimes (ignoring for the sake of simplicity issues of time preference), he'd end up with a higher effective tax rate than I would, because my income is spread out over more years, putting me in a lower bracket. The more progressive the income tax, the less incentive there is to go into medicine.

This also applies to any career that involves working long hours in exchange for more money. Suppose that by working 80 hours per week instead of 40, you can double your salary. But with a progressive tax rate, you take home significantly less than twice as much money, which pushes people away from careers that require long hours. I was considering going to law school at one point, and this was an actual factor in my decision not to.

Another example: Entrepreneurship. Paul Graham has spoken of starting a new company as a way of compressing your work life: Start a company, work like crazy for a period of several years, sell it, and you're set for life. Steeply progressive income taxation makes this less attractive, because your income spikes through the roof the year you sell the company, putting you in the highest tax bracket. And people don't respond by doing it twice to make up the difference--once is bad enough. Instead they just keep working for established companies for the safe paycheck with the relatively low tax rate.

Incentives matter. Even if you think you can outsmart the laws of economics in the short run, they'll come back to bite you in the long run.

Share this

Barrier to entry

For example, to become a doctor takes a lot of work. Four years of college, four of medical school, and then another three to ten of on-the-job training, depending on the specialty. After that, they start making a lot of money. They have to, to make up for the fact that they've missed out on 7-14 years of post-college earning potential, and for the 100-hour workweeks residents have to put up with.

This is a barrier to entry, but much of it has moreover long seemed to me to be a contrived barrier to entry, something created in order to be a barrier to entry rather than a natural and needed part of medicine.

True. And I think we should

True. And I think we should probably reduce the barrier to entry to whatever is needed to ensure adequate competence. But as long as we have that barrier to entry, the fact remains that a steeply progressive income tax makes the problem worse.

I'm in the same boat. I

I'm in the same boat. I would love to work long hours while young so that I could work less when I have a family and kids. But I find it's not really worth working the extra hours, since the taxes are so high. It also hurts that since I'm single and do not own a home, I receive virtually no tax breaks.

It gets worse. Even a "flat

It gets worse. Even a "flat tax" is convex since it becomes 0 for negative income. I think some 95% of businesses created never turn a profit, to tax the 5% that do is revolting. Any convex income tax increases risk aversion, since a regular income is less taxed than an irregular one for the same average return.

Random thoughts

Eh. While this is not my field, I have a few suspicions here.

Does the tax impact on selling a business discourage people from starting businesses? First, I believe the US income tax code provides some opportunities for carry-backs and carry-overs from prior tax years. These provisions permit a kind of income averaging process. Second, a business sale is typically taxed at capital gains rates, which are typically less than income tax rates. Thus entrepreneurs are already getting a tax incentive to forgo their day jobs. Third, people can structure the sales of assets – especially businesses – to spread out the income over time, thereby avoiding some of the tax consequences of the sale. Fourth, it's hard to cheat on paying income taxes on your salary; it's too easily documented. In contrast, business owners have many more opportunities to cheat. In other words, the effective tax rate on business owners as a group is less than the nominal rate. Finally, I believe that entrepreneurs are born, not made. While I don't dismiss the idea that marginal tax rates can have some effect on people's choices, I suspect it's much less than a rational analysis would indicate. That's because I suspect most entrepreneurs are not driven by a rational, dispassionate drive to maximize wealth, but rather by personal preferences to blaze their own trails, be their own bosses, realize their own visions, etc. -- in short, by passions. Thus, I don't harbor any huge concerns that our tax code unduly discourages entrepreneurship.

Is the tax rate structure “convex”? Perhaps, but it’s unclear to me how big a problem this is. True, people with no income pay no US income tax; indeed, the inflection point is not at $0 income, but at the amount of the standard deduction. That said, various government subsidies are tied both to income and to other life circumstances. In this sense, the curve correlating income to wealth transfer may well extend relatively smoothly below the $0 income point. Below a certain income, the net transfer flows from government to the individual rather than the other way around.

If you’re really concerned about the consequences of high marginal tax rates, arguably you’d be most concerned about the poor. People on various government programs face the steepest marginal tax rates because they’re at jeopardy of losing eligibility for government income/health/housing benefits when their income exceeds certain thresholds. The Earned Income Tax Credit ameliorates this problem to some extent, but it’s only a partial solution.

I suspect the issues raised here are most directly remedied by a consumption tax, or perhaps the X Tax proposed by the late David Bradford, principle architect of the 1986 Reagan tax reform. While some people argue that a consumption tax is regressive, Bradford noted that it needn’t be. Conceptually you could retain our current tax code but merely add provisions for an IRA into which a taxpayer may make tax-deductible deposits, and from which a taxpayer may make taxable withdraws, without limit. Such a system would permit people who sell businesses to deposit the money tax-free into such an IRA, and pay taxes on it only upon withdrawal.

Does the tax impact on

Does the tax impact on selling a business discourage people from starting businesses? First, I believe the US income tax code provides some opportunities for carry-backs and carry-overs from prior tax years. These provisions permit a kind of income averaging process.

It does alleviate the problem, but they're definitely not full carry-backs and carry-overs. For capital losses / capital gains for example, the yearly limit is a puny $3,000. Want to go Taleb on the market and buy out of the money puts ? Think again... if you believe the event will pay n:1, you need odds at least (n*t):1 where t is the average tax rate on your gain, which implies a gross mispricing. The more the bet is skewed, the more taxes skew you.

Second, a business sale is typically taxed at capital gains rates, which are typically less than income tax rates. Thus entrepreneurs are already getting a tax incentive to forgo their day jobs.

What about corporate tax ? Besides, even if the tax is slightly lower, it pales in front of the 1:20 bet of starting a business. Overall there are less business created.

Third, people can structure the sales of assets – especially businesses – to spread out the income over time, thereby avoiding some of the tax consequences of the sale.

Little problem, it's illegal.

That's because I suspect most entrepreneurs are not driven by a rational, dispassionate drive to maximize wealth, but rather by personal preferences to blaze their own trails, be their own bosses, realize their own visions, etc.

Someone might start many businesses before one succeeds. There's a limited amount of resources one has to invest in this process. If the business owner could offset the losses of this previous business into the new one, he would avoid many scenarios where the business turns some profit but not enough and fails to grow. Thus there would be more businesses, even considering the psychological factor that you mention.

In this sense, the curve correlating income to wealth transfer may well extend relatively smoothly below the $0 income point. Below a certain income, the net transfer flows from government to the individual rather than the other way around.

Doesn't matter, still convex.

If you’re really concerned about the consequences of high marginal tax rates, arguably you’d be most concerned about the poor. People on various government programs face the steepest marginal tax rates because they’re at jeopardy of losing eligibility for government income/health/housing benefits when their income exceeds certain thresholds.

The big difference is that they don't own what they "lose". But poverty traps are a big problem. In France there is the RMI, roughly $700 a month and other benefits. If you work a little and win, say $100 a month, your RMI is decreased by $100... 100% marginal "tax" rate ftw.

Second, a business sale is

Second, a business sale is typically taxed at capital gains rates, which are typically less than income tax rates.

I'm not 100% sure of this, but I think that to get the capital gains rate your business needs to be incorporated. But if it's incorporated, you have to pay income tax on the profits. On the other hand, if you don't have any profits because you reinvested everything into expansion, maybe you could get away with just paying the 15% rate. Arthur would probably know.

But in another sense it's a moot point anyway, because most of the people who support a steeply progressive income tax also support higher taxes on capital gains.

Is the tax rate structure “convex”?

What does "convex" mean in this context?

If you’re really concerned about the consequences of high marginal tax rates, arguably you’d be most concerned about the poor.

True. They face fairly high marginal rates, and may also have the highest aversion to labor. On the other hand, they just aren't that productive even when they do work. IIRC, the bottom 50% of income earners make about 13% of the income, so if everyone on the dole went to work, it would be unlikely to increase national income by more than a couple percent.

Regarding the X Tax, I had the same idea a couple of years ago. However, exempting all savings from taxation would result in a huge short-term decline in tax revenues unless rates were raised quite a bit. Maybe a Roth-style account would work better: Tax money on the way in, but don't tax gains.

What does "convex" mean in

What does "convex" mean in this context?

The amount of income tax, as a function of income is convex, meaning any linear combination of income will be taxed less than the same linear combination of the amount taxed for each income. Alternatively, it means the marginal tax rate increases with income (progressive tax)