Income Inequality and the Work Ethic

What’s striking about the latest CBO reports showing the ACA leading to millions of fewer jobs is the reason for this. It’s not that these workers are going to be laid off. It’s that they will choose to work less, either to qualify for low-income subsidies or Medicaid or because they will no longer need jobs to get insurance. Some people think this being the basis of the fewer jobs is a good thing a counterpoint to ACA opponents. To me, this is the bigger problem.

And it’s part of a broader issue. There’s a cumulative effect of all the numerous attempts to eliminate income inequality, which start but are by no means limited to the progressive tax rates (and also include income-based reductions to the EIC and the per child tax credit plus any number of other tax-related components) including eligibility for any number of social assistance programs. The ACA is highly progressive as well, including both income-based subsidies for exchange plans as well as features which reduce the cost-sharing components of the plans for lower-income individuals.

People like to harp about the income inequality and make comparisons to other eras and so on, but – from what I’ve seen – they tend to focus on gross income, and thus ignore the real net income-plus-benefits that actually determine living standards. The irony is that efforts to rectify income inequality generally do not focus on gross income, so the “problem” can never be fixed using the standard which is being used to measure it. This guarantees that nominal income inequality can be used in perpetuity to justify an unending list of redistribution schemes, since while these corrective measures impact the lives of the people involved, they don’t impact the criteria used to justify them.

The upshot of all this is that the true marginal “tax” rates are significantly higher than whatever income tax rate is being bandied about, and can act as a brake on motivation to work, as recognized by this CBO report.

Of course, the CBO report is only about the ACA, because they happen to have been tasked with assessing the impact of this particular program. But the same logic applies to all other progressive programs.

[Note: this is not to suggest that all progressive taxes or social assistance programs should be eliminated, of course. Only to say that that the progressive impact of all of them needs to be viewed as a whole, and the impact of this on productivity and work-motivation considered, and I don’t think this is currently being done.]

The new CBO report on Obamacare is front and center on every news site right now. I tend to take such things with a grain of salt. The CBO is non-partisan but it still churns out estimates based on particular assumptions that may be wrong. That said, of course, defenders of the ACA have always been willing to assume that CBO numbers are correct when it serves their purposes.

Your larger point is entirely correct. All of the figures that get tossed around in a debate about income inequality use measures of “income” that ignore government handouts. Contrary to what many people seem to think, handouts have been going up sharply over the last 10 years or so. If we used a measure of income that included those handouts, the results would surely look quite different.

I don’t believe that is true. Here is one cite that shows government transfers have a very limited effect on income inequality. This was in 2005, but even if you double the effect since then you’re still looking at a pretty small effect. Cite:

The data I can find from 2010 puts bottom quintile after-tax income at 3.3% in 2010. Cite: (warning: PDF)

I guess I’d like to see a cite that existing transfers make the distribution look “quite different”. Or a cite that typical reporting on this issue doesn’t include transfer payments - all indications I can find is that they typically do (at least SS, EITC, and things like SNAP - Medicaid benefits are perhaps excluded, and none of the reports include Obamacare subsidies yet, obviously).

It’s not clear from reading your cite what exactly is included in “government transfers”. The source spends some time on defining things but isn’t clear on which are ultimately included in the adjustments. In addition, there are various hefty sources of social assistance which I did not see mentioned altogether, e.g. rent subsidies, college tuition assistance, utility subsidies etc.

But in any event, according to that source, the bottom 20%'s share of income doubled after including transfer payments, which I would think is a substantial difference.

I do believe the CBO numbers are rather comprehensive. Census numbers are less all-encompassing. I don’t think utility and housing subsidies are a significant factor here, but I’m willing to be corrected. There just doesn’t seem to be enough money there (compared to things like EITC) to move the needle.

Well, that doubling is from 3% to 6%. Significance is obviously in the eye of the beholder.

I am sympathetic to the OP arguments regarding effective marginal tax rates, particularly around the phase-out limits for EITC. The ACA may exacerbate this somewhat, but since the subsidies run up to 400% of poverty level, it’s a bit unclear.

It may also have a bigger drag on labor participation beyond just the marginal tax rate effects. Having health insurance is potentially a large psychological motivator for employment, beyond its monetary value. I could see people that don’t really want to work (e.g. second parents that would rather provide child care) be more willing to forgo the additional income knowing that affordable health care was available.

I’m intrigued by some ideas I’ve heard lately (I think it regarding Unemployment Insurance) of just giving lump-sum payouts regardless of whether you get a new job or not. That way you have no disincentive to find work but still have the safety net when you lose your job. Not unlike a negative income tax, I guess.

The most commonly cited data in the debate is the Picketty-Saez data which uses pre-transfer tax units to measure inequality. Using that data the median income for the middle class has gone up 3.2% in the last thirty years. However, if you make the data households, account for household size, after taxes and transfers, and include non-taxed benefits, the median income for the middle class has gone up 36.7% in the last thirty years. So using a measure of income that is post transfer and taxes can have a large effect on the measures of income growth and inequality. For instance if you take 2005 going from pre-tax and transfer data to post, doubles the percentage of national income taken by the bottom 20% and reduces the amount the top 1% receives by one fifth.

According to this, the bottom 20% earn around $10,434 a year. Doubling that to $21,000 a year is still very poor.

So, if your take home is $740 a month [ (10434/12) * 0.85) ] and you are getting the whole suite of assistances (Rent, utility, food, etc) you might get a total compensation package of $840 a month - and most of that is probably rent assistance.

You should also note that the second class, Middle 60%, got 6% of that redistribution, while the bottom 20% got 3%. Why are the not-poorest people getting money? Well…

According to Stephen Miller’s numbers at the rebuttal of this article, $1 trillion is spent each year, including medicaid, in state and federal dollars to service roughly 110 million households. That makes the average cost per household $9,090.91 per year (which includes Medicaid.) Almost half of that is medical spending, so an average benefit would be around $5,000 (45% of 9090.01). That’s $416.67 per month in benefits that can be spent at full or partial discretion of the receiver.

So, if we take a poor rule of thumb of averages and say that double of that average and $1 is the range, some households will be getting upto $833.34 per month, which very closely matches that doubling of income we saw above.

The above ignores all the people who work solely to gain access to insurance. Were it not for the healthcare-job connection that exists in the US, these people wouldn’t have chosen to work in the first place. Is it a societal good to limit such things as “health” only for those who are “employed”?

That’s the moral dysfunction that we must break, and all the CBO whining pretty much ignores this.

The problem with Mankiw’s assessment in this discussion is that if you add in tax breaks and transfers, what happens when those tax breaks and transfers go away? People are concerned with employment sector figures for a reason:

One side wants to eliminate transfers completely (without a plan to make up the lost income in the private sector).

the Other side wants to increase transfers to help the poorer and middle class people get ahead.

Query: If we measure recipients’ income after receipt of assistance, should we measure the income of those in higher quintiles after taxes, to be consistent?

I don’t know. There may be ways to structure transfers in such a way as to reduce disincentives for achievement/work/etc while also providing a robust safety net and rungs up the social ladder. SS is one such transfer program that is relatively mild in its disincentives, since the benefits run pretty high up the income ladder. It’s still pretty progressive, but it also rewards work.

Examining areas where effective marginal tax rates go up near 100% is a decent place to start, I’d think. The one analysis I’ve seen about this (which I can’t seem to find right now) pointed to the range between 100% and 200% of FPL as being a spot where this can happen. That was pre-ACA.

An example of a change that might reduce the effective marginal tax rate for low-income workers would be making the EITC run up to a higher income threshold with no phase-out until you were comfortably middle class. This would obviously cost money, but an analysis of whether the cost is outweighed by removing a disincentive to work would be interesting.

The EITC actually provides a negative marginal income tax rate at the very bottom end, IIRC, providing a huge incentive to work.

I’m not sure why you would say that. I know a lot of people who receive rental assistance and it’s substantial (there are long waiting lists and people can wait for years for a slot, but once they get one it’s gold). But the point here is that there are so many income-based subsidies that the cumulative effect is greater than would appear from looking at any one.

You’re saying that 6% is still a low number. But as a change in ratio it’s a pretty big change.

Why are you multiplying by .85? And why do you assume that rent, utility and food etc. are only $100 a month?

Partially that might be because there are 3 times as many people in the middle 60% as in the bottom 20%.

Ignores? I wrote (and you quoted): “or because they will no longer need jobs to get insurance”.

IMO that depends on whether they’re capable of work. If they’re capable, then they should have to work for their health benefits rather than have other people work even harder to support them.

Of course.

[Perhaps I’m not understanding you here.]

OK, rental subsidies are a bit larger than I thought ($35B in 2012). EITC, by comparison, distributed $56B in 2012. Fortunately, the numbers most used do in fact account for these subsidies when considering income inequality.

I’m not sure that the ratio change is really what matters in this particular calculation. The top 1% provided almost all of that change (going from 19% of income to 15%) - would you say that they only had a 21% reduction (4/19) compared to a 100% increase (3/3) in the bottom quintile? That seems like bad analysis to me.

It’s a bit more complicated, because rental subsidies are concentrated at lower income levels than EITC.

When I was following the program more closely, the rule (at least in NJ) was that you paid 30% of your income for rent and HUD paid anything above that. Which meant you had an effective 30% marginal “tax” rate on income, over and above any taxes, EITC, or loss of other social programs.

It matters because these ratios are frequently used to sell the idea of income redistribution.

Sure, some people may work fewer hours to keep subsidies ,but it is not unheard of for both halves of a couple to work , one for a higher income and the other to obtain various benefits even though the single , higher income is sufficient. Insurance has been one of those benefits , along with pensions, free or reduced tuition, terrific employee discounts, etc. In some cases this was because employer-sponsored plans end up being less expensive than individual plans and in others because of pre-existing condition exclusions on individual policies. I’ve known more than one woman who kept working for the benefits even though her husband made enough in self-employment to support the whole family and she would have preferred to be a SAHM . I suspect there are also a fair number of people who kept jobs instead of striking out on their own because of insurance issues.

OP is correct that subsidizing health-care (and other expenses) for the poor creates a disincentive to work. The simple solution, which is espoused by many economists on both the “left” and the "right" is to subsidize health care (and perhaps other expenses) for ALL citizens, not just the poor and unemployed.

This, e.g. government-paid health care, is of course adopted with great success in many European countries. Unfortunately it is *impossible *in the U.S. of America becauseTaxes would have to be increased.

Thanks for the details on that. If those numbers are correct for that program, then it’s clearly a disincentive to work. It seems to me that a flat contribution from the recipient (perhaps with tiers based on income level, although that creates rate spikes at the thresholds) would be a smarter way to administer that program: it wouldn’t require income verification and it wouldn’t create that high effective marginal rate.

I think the comment was made up-thread (or in another thread) - pointing out these potential problems is great, but not worth much without stating some alternatives. Just taking away housing subsidies removes the disincentives, but I’m not sure you would want to argue that their removal without replacement would be a net positive for society (or maybe you would?).

As I understand it, you’re pointing out that some people will be self-supporting in any event, but only keep a job because the market for individual health insurance pre-ACA was so thin.

That’s a fair point.

Of course, if those taxes were increased in a progressive manner, then this too would function as a deterrent to working.

If you don’t have tiers then you have a massive loss at the point where you get disqualified. If you do, then you have smaller losses at the breakpoints, as you note. Either way, at some point you’re going to run into a guy losing money because he got a raise or worked extra hours or went to school and got a better job.

I wouldn’t know how to solve everything. It’s very complicated, and one problem of starting massive social programs - or even a massive amount of smaller ones - is that you change people’s lifestyles and mentalities and situations in ways that can’t easily be undone.

My point in this thread was more limited. Only that when contemplating any specific progressive program or tax, the current situation needs to be viewed based on the net income/benefits after the cumulative effect of the multitude of different taxes and social programs. It does not seem that this is currently being done.

I don’t have much to add except to not that this is an excellent point and something I have wondered about to. To be a truly accurate picture, income for these purposes should include all subsidies on one side and subtract out all liabilities on the other. I believe a truly accurate picture will show that income equality isn’t nearly as great in many or most cases as we are led to believe and that leads to a further distortion based on social policy.

I know that it is complicated and depends on the specifics in question but I noticed a long time ago that it is sometimes financially advantageous to be poor rather than lower to middle-middle class and sometimes even above that. One example would be a poor family in subsidized housing on food stamps with two smart high school kids who want to go to college. Many colleges, especially the best ones, will fall all over themselves to give the kids a full ride scholarship because of their parent’s low income.

The overall subsidized lifestyle of the low income family may not be significantly worse than a somewhat wealthier family down the street who pays their own bills but the very poor family’s low income triggers a chain reaction of subsidizes that can be worth well over $100K a year or even more especially when you are talking about scholarships combined with everything else. Meanwhile, the wealthier families around them can’t even begin to afford to send their kids to any decent college because they don’t have any free cash and yet they make too much money to qualify for the truly generous subsidies. The poor family got a large reward just for being broke and I don’t think that is a good thing overall. Please note that I was that ‘poor’ kid who got a full scholarship because my stepfather went bankrupt shortly after he married my mother and I still don’t think it was fair.

That is just one example of many possible scenarios. It is hard to quantify because federal, state and local subsidies are such a patchwork tapestry in the U.S. but I believe it is a very valid point that all of them should be included when determining someone’s true income and when debating income inequality in general.

This is straight-up wrong. I’m surprised nobody called you on this yet, so let me be the first to point out that no, the latest CBO report does not show the ACA leading to millions of fewer jobs. It shows the ACA leading to a reduction of hours for many of the equivalency to millions of fewer jobs. The big difference, of course, being that this is a voluntary supply-side movement - these businesses whose workers now have less hours still need the same labor done. What’s more, the problem with Obamacare is not that it’s your standard progressive income scheme. It’s not just “get more money and we will take more from you” (something which almost certainly does not have a significant effect at current rates), but rather that it’s sort of a notch effect - “if you earn above this point, you lose a massive benefit”. That kind of thing is harmful, because it creates a gap in the correlation between income taken in and effective earnings, a gap which can be quite hard to bridge.

To what degree, though? I don’t buy this whole premise, that marginal tax rates are a deterrent to working. It’s like when Bill O’Reilly complains about higher taxes making him less incentivized to work… Oh, okay Bill, so you’d rather earn no money, rather than slightly less money? :rolleyes: The fact is that discounting “notched” programs, which are entirely a problem that needs addressing, I just don’t see the mechanism where tax rates below even 90ish percent cause people to lose interest in earning a living. Let alone the far more realistic 17-30% we’re talking about for those in the middle class. You have the choice between taking home more money (which is taxed), or not.

See, under no circumstances is this actually necessary. Well, okay, one circumstance - when politics sticks its dick in it.