A young adult UBI?

Gareth Morgan's party proposes $3.4b to go to everyone aged 18-23 as $10k after tax transfer – a limited UBI.

I had a short chat with The Project about it yesterday; the logistics didn't work out for a longer chat as I was out to Christchurch to help launch an excellent new book on smart water markets - more on that another time. I'd put together a few notes in case I was to have had a longer chat; I'll share those here.

The bulk of Morgan's proposal would be funded by cancelling National's tax package, with minor bits coming from forecast future surpluses ($400m), canning student allowances and student loan living costs for folks in that age cohort (maybe $267m they think), and job-seeker support for those in that age cohort.

I don't know how this interacts with WFF and consequent fiscal effect. Some of the benefit could be clawed back - with potentially lower fiscal cost.

The bulk of the cost comes from cancelling National’s tax package. That package pushed out the income tax thresholds for the two lower tax boundaries. So providing funds to those 18-23 is at the expense of reduced taxes for every other cohort.

National’s tax package bumped up the accommodation benefit for students and hiked the accommodation supplement. In current rental markets, that programme would mostly subsidise landlords rather than help tenants; just giving that money as cash transfer to 18-23 year olds may not be all that bad.

More generally, there are two basic ways of trying to provide income support. Targeted programmes, like those that the government currently runs, and like those it will be further developing as part of the investment approach, seek to direct funds to particular sorts of need. They get messy and complicated very quickly as necessary part of targeting, and the rules can often feel perverse. If you want to make sure that kids in households with the least support get the most help, you need checks around what kinds of support are available in the household – and that’s where all of the monitoring stuff around live-in partners and the like comes in.

These programmes are able to deliver targeted benefits at tolerable cost, focused most closely on areas of greatest identified need. And the Investment Approach will ramp all of that up to direct funds to programmes that do the most good in improving lives, as measured by reduced reliance on benefits. Note that the object there isn’t the reduced reliance on benefits but that it’s a signal of other things having gone wrong.

A UBI is at the opposite end of the scale. It provides blanket payments to everybody regardless of need. The UBI forgoes targeting in favour of simplicity, but at the expense of high cost. So while a UBI would reduce the high EMTRs facing a lot of people on multiple benefits who are working 20-30 hours per week, where combined clawback rates can mean that workers only keep 10 cents or less from each dollar earned (in some cases), it is at the expense of higher EMTRs for all other earners. And then the net effect depends on whether you do more good by reducing large perverse incentives for a small group of people, or by avoiding (relatively) smaller perverse incentives for a much larger group of people.

TOP is right to point to the unfairness of some of the support provided to students that is not provided to others starting out in the workforce. They propose taking away some of the extra support provided to tertiary students (though fall short of re-introducing interest on student loans, which they should have), but apply the savings to a blanket payment to everyone aged 18-23 regardless of need.

And where they've maintained benefit payments above $10k for those currently in receipt of benefit packages over $10k, they've also maintained some of the costly hoop-jumping (and high EMTRs) that are part of the costs of the current system.

As for incentive effects and work, here's a recent evaluation of what happened in Manitoba's Mincome experiment.

Thus, Figure 5 graphs overall trajectories in order to get a general picture of subgroup trends. Subgroups are displayed as baseline and study period averages for ease of presentation, and treatment effects are shown in parentheses in Figure 5. Among the most consequential, Mincome’s average treatment effect (the difference between changes in Dauphin and changes in the Manitoba control) for singles is a 16.2 percentage point fall in household participation in the labor market. Among young people there is a similarly large treatment effect, at 18.6 percentage points. Dual-headed households appear less sensitive to Mincome. For this group, the equivalent treatment effect is 7.4 percentage points. Thus, the overall experimental effect on labor market participation is disproportionately driven by changes in young and single-headed households.

So the biggest drop in labour market participation was among youths. 

Dauphin's guaranteed family income started at $19,500: about half of annual household income at the time. Morgan's proposed $10,000 is much lower than that relative to median household income in New Zealand, so corresponding effects on labour market participation would be expected to be lower than those found in Manitoba.

While the authors note that drops in participation would not be large enough to cause problems in overall scheme financing, note that Mincome wasn't self-financing. Lots of people in Dauphin weren't in the experiment, and the money for the experiment came from overall government revenues. 

That makes it harder to tell what the real effect on participation would be. On the one hand, you might expect people who wanted to be able to drop out of the labour force would be disproportionately willing to participate in the experiment, which would mean the found effect is larger than you might expect for the population overall. On the other hand, you might expect that if everyone faced the kinds of tax rates necessary to fund a UBI scheme, dropping out of work for current workers would look more attractive. In that case, you'd expect real-world effects to be larger than those found in Mincome for payments comparable to those used in Mincome. 

The UBI and the political constraint

I walk through the basics on universal basic incomes over at The Spinoff. After explaining Milligan's impossible trinity, I note the political constraint that I think blocks Gareth Morgan's proposed UBI. Morgan's proposed UBI would leave some very poor families worse off:

More importantly, though, while relatively low basic benefits could make the system affordable, they would not be politically stable. There would be pressure to layer a benefits system on top of the UBI, or to increase the UBI. It does not seem plausible that any government would be able to withstand the likely months’ long John Campbell campaign that would, every day, highlight a different family whose benefits were cut under the shift to a UBI. We would quickly have a welfare system layered on top of a UBI, increasing the costs while eroding the UBI’s benefits.

I could imagine pushing a button for a low UBI replacing existing income support programmes, combined with generous tax credits for contributions to charities that work to plug remaining gaps. But that button doesn't exist, and even if it did, it wouldn't be politically stable. You would have, every night, John Campbell featuring some family that has been made worse off. Because that is what John Campbell does, and because there is a market for those stories. Even if charitable support programmes were ramping up and would be more effective in the long run, the political constraint would bind. And then we wind up with UBI plus existing welfare.

Over at The Sandpit, I note a few additional problems that John Gibson raises:

[Gibson] then goes through five rather important potential unintended consequences of income transfer schemes, mostly looking at things in developing countries:

1. Transfers targeted to types of need encourage Tullock-style competing for aid. He notes a programme in Brazil targeting poor families with kids wound up resulting in less weight gain per month for the targeted kids; the families feared that if their kids grew well, the transfers would stop.
2. If being in formal work means you pay taxes to support protection schemes targeting those outside of formal work, then you get distortions towards informal work;
3. Transfers to targeted rural households generate localised inflation that hurts non-targeted households;
4. Programmes can erode existing informal safety nets where people rely on each other and family during tough times. The effect of programmes is then a bit harder to judge where the effective beneficiary is the person who would otherwise be supporting the recipient of aid.
5. Programmes likely affect household composition, and consequently undermine targeting. He gives the example of the expansion of old age pensions in South Africa: adults with low skills wound up moving in with pensioners.

GAI

There's a lot to like about a guaranteed annual income. Or, at least, there would be if it were feasible and affordable. I don't think it can be both.

Let's recap. On Friday, Labour released a discussion document on the GAI. The document is much better than I had expected, laying out reasonably the benefits and costs of that kind of a system, with a lighter thumb on the scales than I might have expected.* 

The main problem remains linked to Kevin Milligan's trilemma.

@EmmMacfarlane Again, the impossible trinity: choose two of:

* low phaseout rate
* Big basic benefit
* same cost as current system.

— Kevin Milligan (@kevinmilligan) February 5, 2016

You can't pay a benefit large enough to leave current beneficiaries no worse off without simultaneously blowing out the budget, or running a very sharp phaseout rate. This isn't rocket science. The beneficiaries in most need get lots of different payments from lots of different systems because they have multiple needs. Set a GAI to replace all of them, and you'll leave the worst off worse off, or you wind up paying everybody the amount of money that we currently pay to those being paid the very most.

And so we come to Labour's document. They note the main benefits of shifting to a GAI: getting rid of the very high effective marginal tax rates affecting the working poor where multiple programmes abate simultaneously, getting rid of the stigma of benefit receipt,** and getting rid of the paternalism inherent in the benefits system.

When they get to the costs side, they note that a GAI would be rather expensive and that a GAI sufficient to fully compensate those beneficiaries in the worst circumstances would be on the order of $22,000 per year. Note that total Core Crown revenue is about $72 billion per year. A GAI of $22,000 per person would exhaust total Core Crown revenue when paid to just over three and a quarter million people. There are 4.6 million people in New Zealand. Again, not rocket science.

So how to square the circle? Their options suggest a lower GAI, but with top-ups for those in more severe need. But that's a problem. The benefit of having a GAI - getting rid of paternalism, stigma, high EMTRs - putting need-based top-ups on top of a GAI brings the problems of the current welfare system back.

Libertarians abroad who favour a GAI have good reasons for doing so. They mostly suggest funding it by cleaning up foreign tax systems - abolishing the kinds of exemptions and loopholes that New Zealand doesn't have. Now that's still a bit of a cheat: those changes are worth doing regardless of a GAI and should be evaluated separately. But NZ's tax code doesn't have free lunches baked into it - thankfully.

A GAI then winds up paying a fair bit of money to non-working spouses in high income families and to students while potentially making it harder for current beneficiaries in need to access benefits, if the top-ups follow the form of current hardship grants for beneficiaries. And at a not insignificant hike in tax rates.

Treasury's 2010 analysis remains relevant.

Update: And here's the 2001 McLeod Report:

Universal basic income 6.52 Submissions reiterated the case for a universal basic income (UBI).

6.53 A UBI provides a fixed sum to each citizen. Key attractions of this idea are its simple administration and avoiding EMTRs from abatement. There are also philosophical arguments for a UBI (everyone has a right to a basic income; a basic income reflects a return to collective wealth) and against (people have a right to the fruit of their labour).

6.54 There are three practical problems with all UBI proposals, namely:

• a UBI provides people with money (which gives them purchasing power over goods) without supporting the production of the goods to be purchased with the money;
• income distribution: New Zealand has few high-income people and many low income people. Each dollar taxed off the few people at the top of the distribution has to be divided among many people at the bottom. This, in turn, means either the UBI has to be low or the tax rate to fund it has to be very high; and
• churn: The people in the middle of the income distribution pay half their income in tax and receive the same amount back as UBI. Much of the high tax rates of a UBI scheme is required to take money from middle- and high-income people and give it back to them, worsening their incentives without increasing their net income. This means we get the costs of high tax rates without the benefits.

6.55 A UBI has theoretical attractions, but the high tax rates required to fund it and the incentive effects of the payment make it impractical.

* I would link it, but Chrome really really doesn't want to open Labour's page due to an https issue.

** I'm not entirely convinced this is an unmitigated benefit. 

Guaranteed Minimum Income

TANSTAAFL applies, as always.

Matt Zwolinski makes the libertarian case for a guaranteed annual income over at the BHL blog. It's not a bad case. Here's Matt:

The first of these posts elaborated on the (quasi-Nozickian) argument that a BIG could serve as a kind of rough-and-ready compensation for past injustice. David Friedman and David Henderson both took issue with this argument as I articulated it in my original essay. But I’m not convinced. The federal government was directly responsible and/or culpably complicit in the commission of a long series of gross injustices, and many currently existing Americans continue to suffer the effects of those injustices. The government owes those who were harmed by its wrongdoing some form of redress, and I think there are plausible grounds for using a BIG to make that redress. I haven’t seen any published responses to this essay yet, but I’d be interested in talking more about the issues of historical redress, collective responsibility, and second-best theory that it raises.

My second post tried to explain why Friedrich Hayek supported a basic income. Or, perhaps more accurately, it tried to develop an argument based on Hayekian considerations. It’s not an argument that Hayek actually made. But it’s an argument that’s rooted in the quasi-republican account of freedom and coercion that he developed in The Constitution of Liberty, and I think it’s an argument he probably would have endorsed. Essentially, the argument is that a basic income is necessary to keep people out of the kind of poverty that could render them vulnerable to coercion by employers and others with economic power. I’m sure a lot of libertarians will bristle at the notions of “coercion” and “freedom” employed in the argument, but I think that there’s a lot to be said on their behalf, and that they’re worth taking seriously.

He also has a great links roundup of others' commentary; David Friedman's critique is particularly good.

The basic problem with guaranteed income schemes is that they cannot really replace the welfare state. It would be great if they could, because we could then do away with a whole pile of messiness around really high effective marginal tax rates when abatement starts kicking in on a bunch of means-tested policies, and a whole pile of transactions costs around assessing eligibility for different benefits.

Think of it this way. Imagine the level of support currently granted to somebody who's really badly off: a bunch of disability-related benefits, income support, child-related support, housing support and the like. The total value of that bundle is going to be large. If you set up a GAI to make sure this person is no worse off, the average and marginal tax rates needed to support paying everybody that amount would be crippling. If you instead set up a GAI to pay everybody a basic income and have other support schemes on top of that for those with particular needs, you keep much of the current welfare apparatus but add in the substantial hit to marginal tax rates of giving everybody the GAI payment.

I don't think the basic math's changed much since I wrote about it in 2011; Treasury had then reckoned we'd need a flat personal tax rate of between 50% and 56% to make it work. Again, we're faced with Milligan's Impossibility: In any GAI, you can only pick two of the following three:

1. low tax rate
2. high benefit
3. balanced budget

The University of Manitoba's Evelyn Forget has done some nice work on a guaranteed minimum income scheme trialed in Dauphin a few decades ago. Straight cash transfers to people in Dauphin from people mostly outside of Dauphin did a lot of good for a lot of people in Dauphin. I wonder what it would have looked like if Dauphin's higher income earners and businesses had to cover the programme's costs.

The impossible trifecta

Kevin Milligan says a Guaranteed Annual Income cannot simultaneously satisfy three goals. Instead, you have to pick two among the following:

1. low tax rate
2. high benefit
3. balanced budget*

Treasury here in New Zealand modelled a GAI for New Zealand on the request of the Welfare Working Group (HT: Lindsay Mitchell). What did they find? A GMI paying $300 per week - the mean benefit income among those on benefits - would cost $44.5 billion, or $52.6 billion if we extended it to superannuitants as a replacement for NZ Super. The former could be covered by a flat personal income tax rate of 45.4%; the latter, 48.6%. But full fiscal neutrality would require tax rates of 50.6% and 54.4% - the lower tax rates would be just enough to cover the transfers, but income tax revenues are currently also used to fund more than just transfers.

If we recognize that most parents are beneficiaries via Working for Families and compensate them for the loss of our version of EITC with a $86 per child per week payment, we get a $57.1 billion fiscal cost and a personal tax rate of 50% (or 55.7% for fiscal neutrality).

And, even this rather expensive system leaves the worst off worse off, as it kinda has to. Treasury notes:

Although the Gini coefficient improves under all models, many beneficiaries (including the disabled, carers and sole parents) currently receive more than $300 per week and would be made financially worse off under a GMI scheme. Therefore the GMIs considered could distribute money away from those most in need of government assistance and toward those who have choices and opportunities but choose not to work.

Treasury also warned about potential adverse labour supply responses to the necessary personal tax rates. And, the induced gap between company and personal tax rates would increase IRD's enforcement costs.

Treasury concludes by reiterating Milligan's impossibility:

From the international examples it is apparent that the more equal a society is in the beginning, the lower the returns to a GMI scheme. That is, a New Zealand specific GMI would either be at a level of income too low to reduce poverty, or a level of income that is high enough to reduce poverty but is therefore expensive and hence distortionary through higher tax rates. 

I don't think Gareth Morgan's Big Kahuna scheme is able to escape the trifecta by imposing new taxes on capital or land. Why? Because those aren't free lunches either. If it's worth having a land tax, it's worth doing it regardless of whether we have a GAI. So in the first stage we set the optimal tax structure - and I'm completely unconvinced that a capital gains tax is all that hot an idea anyway (see Seamus's posts here here and here.) But whatever the optimal tax structure, we implement it in stage one. Then, we still have to increase all the tax rates by enough to pay for a GAI if we're going to have a GAI. And the impossibility reasserts itself.

Morgan squares things with a cheaper GAI paying $11k instead of Treasury's $15k. But I have a hard time believing that's a stable political equilibrium. Could NZ politicos really avoid the temptation of adding targeted benefits for the many folks currently on benefits totalling well over $15k? If not, how quickly do we wind up having a GAI on top of a targeted benefit system?

* Update: Milligan's updated this in a more recent tweet. I think there are a few possible impossibilities here.

@EmmMacfarlane Again, the impossible trinity: choose two of:

* low phaseout rate
* Big basic benefit
* same cost as current system.

— Kevin Milligan (@kevinmilligan) February 5, 2016