That still has the effect of a (probably high) marginal tax and disincentivizes increasing one's income just the same. Why take night classes or work harder at your job if the $100/week raise comes with $80/week less in rent subsidies + taxes? (Example figures; you can change it around but most realistic situations are going to look similar.)
It's true that the price of receiving a subsidy at all is that any money you make is going to be effectively taxed at a higher rate than it would have been otherwise. I think people can live with that as long as the effective tax rate doesn't get too high. Not sure what "too high" would be for most people, but I would guess that as long as the total effective incremental tax rate is never over 60% — meaning that after the reduction of the subsidy plus income tax, one would never keep less than $.40 of one's next dollar in earned income — that would probably be good enough. Yes, there will still be people who will find that a barrier, but at least there won't be this point where your next dollar of earned income will be effectively taxed at 50,000%.
The extreme version is UBI, where everyone receives the full subsidy unconditionally, and the effective incremental tax rate is always just the income tax rate. I'm not sure that's a bad idea, but we could get rid of the current cliff much more easily than we could institute a full-blown UBI.
>It's true that the price of receiving a subsidy at all is that any money you make is going to be effectively taxed at a higher rate than it would have been otherwise.
I wasn't referring to that; I was referring to the decreasing-subsidy-as-income-goes-up as being effectively a marginal tax. (I think the rest of your comment is recognizing it as one.)
>I would guess that as long as the total effective incremental tax rate is never over 60% — meaning that after the reduction of the subsidy plus income tax, one would never keep less than $.40 of one's next dollar in earned income — that would probably be good enough. Yes, there will still be people who will find that a barrier, but at least there won't be this point where your next dollar of earned income will be effectively taxed at 50,000%.
Even a 60% marginal tax is very strongly disincentivizing and makes it difficult to justify most any feasible long-term plan for increasing one's productivity from a poverty level.
>The extreme version is UBI, where everyone receives the full subsidy unconditionally, and the effective incremental tax rate is always just the income tax rate. I'm not sure that's a bad idea, but we could get rid of the current cliff much more easily than we could institute a full-blown UBI.
Right, that's the major upside of UBI. The downside is that it's really expensive (in terms of tax rate relative to benefits) and can't be concentrated on people who might need more help (temporary or otherwise) than others. And if you combine the two, then it's an even more punishing tax with its own effects.
> Even a 60% marginal tax is very strongly disincentivizing and makes it difficult to justify most any feasible long-term plan for increasing one's productivity from a poverty level.
Are you aware of any research that bears on this question? Right now we just seem to have dueling intuitions.
Anyway, I already agreed that some people will find it disincentivizing. But again, the disincentive can't be anything like the massive one we currently have with the cliff. Do you really not think that what I'm proposing would be an improvement over the status quo?
Okay, let me give you the details of intuition then: the typical person in poverty might optimistically hope to go from $30k/year to $60k/year. But the typical such person has a high time preference and is only able to defer earning income under very difficult terms (e.g. high interest rates). But at a 60% effective tax on making that transition and given the long time it takes to bring it about, you need someone with an unusually low time preference and high luck for that to look worthwhile.
(You would be correct that you could have a poverty program that directly targets the high impediments that exist in the short term e.g. getting childcare for when you have the night classes, low-interest loans to cover short term expenses, etc. But those would also be different policies from simply giving stuff at a lower-effective-marginal-rate than the step function does, and it would be adding the kinds of incentive effects that I have been emphasizing as more important.)
If you model everyone as being the go-getter high-potential individual that you are, you'll have a different intuition on it; I don't think that model is accurate on average.
Just to be clear, I don't necessarily oppose those other programs, and I never suggested that getting rid of the step function was the only thing that should be done, though I continue to believe that that in itself might be a very substantial improvement. At the very least, any welfare reform bill should clearly include that as a key component.
"Even a 60% marginal tax is very strongly disincentivizing"
Citation needed. What I've read is that even an 80% marginal tax does not create strong disincentives.
Think of it this way. Most people spend 95-105% of every dollar they earn, and very few of those dollars are very elastic. The vast majority of those dollars are already allocated for taxes, rent/mortgage, food, car payments, et cetera and only a couple of pennies of every dollar made is discretionary. But if you're able to pick up an extra shift, the extra dollars you make from that are purely discretionary. So even 20 cents on the dollar is a windfall.
I'm not sure what literature is out there. My own sense though is:
1. I personally am very attuned to marginal tax rates and opportunity costs. e.g.
a. I won't pick up the extra shift because I'd rather be doing something else than earning $0.2/dollar. (This happened to me in my lower-income days as a student. Due to X hours/week of being a TA giving a tuition waiver, I would never work more than X hours due to the marginal income beyond X hours being so much lower)
b. This happens to me today where I'll take long breaks between jobs I might otherwise not take. I take longer breaks than I would if we had a flat tax, because my loss of income is (1-marginal_rate) * salary, rather than the significantly higher (1-effective_rate) * salary. Similar thing applies in windfall years (e.g. IPO generating large taxable incomes), where I'll opportunistically take months off.
2. I've generally found others to be much less in tune to marginal rates, etc. So my hunch is that marginal rates are not that disincentiving as GP might suggest.
If you want to abolish the marginal tax, then I would start by looking at tax reform. If you think the marginal tax imposed by a given welfare system is too high, you can just tweak the numbers.
I meant that it's effectively a high marginal tax, when you decrease the housing subsidy as the recipient earns more income.
That is, if every $100/month gain in wages comes with a $30/month decrease in subsidy, that's effectively a 30% marginal tax (on top of all their other taxes).
(And I don't know what you mean by "just tweak the numbers"; the more generous benefits you want, the more you have to disincentivize work; no amount of "tweaking" avoids the core tradeoff.)
What do you want the marginal rate to be? 1%? Just have a $1/month decrease in the subsidy for every $100/month increase in wages.
Personally, I would prefer if all welfare was unconditional and we did all of this tweaking through the explicit tax code, but that is an implementation detail.
>The more generous benefits you want, the more you have to disincentivize work;
What disincentives work is the marginal decrease in benefits. Assuming you have adequate funding, there is no need to increase the rate of marginal decrease. If you don't have adequate funding, then you will need to fund the increased benefits to the not-as-poor by decreasing benefits to the poorer.
There is, admitadly, a decreasing marginal value of money that does behave like you are describing, but its effect is dwarfed at the level of income we are talking about.
If you could go from $200/week salary + $200/week welfare to $300/week salary + $120/week welfare, it would be in your interest to do so, as the total is more in the second scenario.
