Scott Sumner  

Experiments in a Guaranteed Annual Income

PRINT
ISIS and Reproach... Hell Nyet, We Won't Go...

Ever since the 1970s, I've favored wage subsidies for low wage jobs, and I've opposed the minimum wage and the guaranteed annual income (GAI.) It seems to me that this article in The Economist has some bearing on the debate over the GAI:

ON A rainy weekday afternoon, Mike Justice pushes his two-year-old son in a pram up a hill on the Siletz Reservation, a desolate, wooded area along the coast of Oregon. Although there are jobs at the nearby casino, Mr Justice, a member of the nearly 5,000-strong Siletz tribe, is unemployed. He and his girlfriend Jamie, a recovering drug addict, live off her welfare payments of a few hundred dollars a month, plus the roughly $1,200 he receives annually in "per capita payments", cash the tribe distributes each year from its casino profits. That puts the family of three below the poverty line.

It is not ideal, Mr Justice admits, but he says it is better than pouring hours into a casino job that pays minimum wage and barely covers the cost of commuting. Some 13% of Mr Justice's tribe work at the Chinook Winds Casino, including his mother, but it does not appeal to him. The casino lies an hour away down a long, windy road. He has no car, and the shuttle bus runs only a few times a day. "Once you get off your shift, you may have to wait three hours for the shuttle, and then spend another hour on the road," he says. "For me, it's just not worth it."

. . .

Small tribes with land close to big cities have done well. Yet a new study in the American Indian Law Journal suggests that growing tribal gaming revenues can make poverty worse. The study looks at two dozen tribes in the Pacific north-west between 2000 and 2010. During that time, casinos owned by those tribes doubled their total annual take in real terms, to $2.7 billion. Yet the tribes' mean poverty rate rose from 25% to 29%. Some tribes did worse: among the Siletz poverty jumped from 21.1% to 37.8%.

Experts offer several explanations. Drug and alcohol abuse are rampant on reservations, so many tribal members find it hard to hold down a steady job. Poor health is another problem: Native Americans have high rates of obesity and diabetes, which are often aggravated by a lack of good medical care.

. . .

But the biggest problem may be the way casino profits are sometimes disbursed. Per capita payments have grown as gaming revenues have risen. "These payments can be destructive because the more generous they become, the more people fall into the trap of not working," says Ron Whitener, a law professor, tribal judge and a member of the Squaxin Island Tribe in Washington state. Of the 17 tribes in the study that handed casino profits directly to members, ten saw their poverty rates rise. Of the seven tribes that did not, only two saw such an increase (see chart).

Per capita payments range from as little as a few hundred dollars a year to more than $100,000. In some tribes, members receive 18 years of per capita payments in a lump sum when they turn 18. "There are a lot of very successful car dealerships around reservations that make their money off 18-year-old[s]," adds Mr Whitener.

One very small tribe in the study, Jamestown S'Klallam in northern Washington, has eliminated poverty entirely. That tribe does not issue any per capita payments and has used its casino profits to diversify into other businesses, such as harvesting huge molluscs for export to China. Squaxin Island, which reduced its poverty rate from 31.4% in 2000 to 12.4% in 2010, used casino profits to get into cigarette manufacturing about ten years ago. Leaders of the Siletz tribe, by contrast, allot 40% of the casino's net revenues to per capita payments and only 17% towards economic development. Of the tribes surveyed, the Siletz has one of the highest poverty rates.

. . .

Kevin Goodell, a 51-year-old Siletz tribal member, works on the tribe's forestry crew. It had several job openings last year--but no qualified applicants, according to Mr Goodell. He says he tried to get young people interested, but they told him they didn't want to work in the woods. With free housing and health care, "a lot of people have figured out a way to use the system to survive," he says. "Why get a job if you don't need one?"


Yes, why get a job?


Comments and Sharing


CATEGORIES: Labor Market




COMMENTS (35 to date)
david writes:

$1200 annually doesn't seem like a sum that should plausibly be enough to disincentivize work. Per day, it's half an hour at the minimum wage rate ($3~).

Let me suggest that possibly reservations with high levels of alcohol/drug abuse are especially likely to decide to dispense cash payouts.

Floccina writes:

I will have to adjust my priors but...
The goal is to create a welfare system that is efficient in that it does not waste money on overhead and minimizes discouragement of work. One question would be do the likes of Mr Justice work for in family consumption. For example does he garden, hunt, fish do home repair or even work for cash illegally. If so he is working, just not in the official taxed economy. I would hope that he is.
Anyway the system that we have now is so inefficient the GAI might still be better. I know from other posts that you prefer a work subsidy (like the EITC) but I think if you make the work subsidy big enough sham companies can be created that require no work and so you end up with a GAI anyway but with money wasted on enforcement and with the EITC you have to keep SS and SS disability, though you can dump the min wage. Of course replacing TANF and SNAP and some other programs with a bigger EITC with a weekly payout would be much better than what we do now.
PS we know that Mr Justice thinks he has enough money because he has a chance to work so Democrats cannot trot him out as showing how terrible we are for not helping him more.

Jim Glass writes:

Jim Manzi a while back had some comments based on studies of a closely related proposal, showing effects of a Negative Income Tax in actual practice when tried, with links to sources. FWIW.

Jay writes:

Am I supposed to feel sorry for Mr Justice or that there's some societal ill there to fix? His choices and plight seem to be all his making. He chooses to live on the reservation but an hour away from the only source of work for which he could get a job readily but chooses not to, and also chooses to support a family he can't afford without making the sacrifices necessary (i.e. moving) to properly provide for them.

amelanchier writes:

These aren't proper experiments in the GAI. Apart from the selection issue brought up by david, there is the fact that these payments are on top of the existing welfare state, with its numerous poverty traps. Milton Friedman's GAI proposal was as a replacement for the welfare state, and as such it would unambiguously improve work incentives: http://www.economonitor.com/dolanecon/2014/01/03/the-economic-case-for-a-universal-basic-income/

Hazel Meade writes:

@david, your math is off.
$1200/year is much less than $3 per hour. More like $0.60 / hour.

$1200/12 = $100/month
$100/160 hours = 62.5 cents per hour.

Hazel Meade writes:

The GAI is going to face major demographic problems if ever implemented. This is due to the fact that people with a lot of free time tend to have more babies. And children that grow up in welfare households are much less likely to be employed as adults. In other words, the GAI is likely to breed more recipients which spells instability. Babies are not randomly and uniformly distributed throughout the population.

Scott Sumner writes:

David, Don't forget they also get welfare.

Floccina, Good points. To reduce fraud I'd run these programs at the local level.

Thanks Jim.

Jay, I try to avoid thinking in terms of villains and victims, but rather as a utilitarian.

Amelanchier, I agree that it's far from a perfect test, but I think it does tell us something about the impact of a GAI on incentives.

Having said that, it's quite possible that a GAI would be better than our current system. My fear is that it would add onto our current system (which is also a problem for wage subsidies.)

Michael Byrnes writes:

I don't think this example is all that useful for evaluating the pros and cons of a GAI.

1. The per capita payment isn't given as a substitute for some or all safety net programs. (Indeed, part of his support was his girlfriend's welfare payment. Oh, and as an aside, why is her status as a recovering drug addict relevant to the topic of this article?) We don't know what other types of support they might be receiving, but given the patchworkiness of our safety net there are probably more benefits).

2. It's not reasonable to focus on the work disincentive effects of a GAI in isolation from those of a wage subsidy system. John Cochrane notes in a post here (
http://johnhcochrane.blogspot.com/2012/11/taxes-and-cliffs.html?m=1) that the disincentive effects of our current safety net are extremely high; some workers face implicit marginal rates that are far higher than those faces by those in the highest income brackets. Ed Dolan has written about this as well. Its fair to argue that a GAI will disincentive work, but you also need to consider the disincentive effects of benefit phaseouts. It's at least conceivable that the latter will outweigh the former, in which case a GAI would provide a net incentive to work (even though many are disincentivized).

3. If some people choose not to work under a GAI, who cares? Won't this be self-limiting? If enough people do this to put a real dent in output, then prices will rise, the real value of the GAI will fall, and the disincentive effects will be weaker.

4. A wage subsidy system is more susceptible to being gamed, in various ways. Of course we could spend money on enforcement, but that will just mean more bureaucracy.

5. If our society really does reach a stage where we are less dependent on physical labor, a GAI seems like a better way to transition there than subsidizing people to dig holes and fill them in again.

JJ writes:

Hazel and Scott,

Is there anyway to socially engineer Welfare with the right incentives that would lead to the least societal problems. Your issue is that GAI would lead to Welfare recipients creating more babies in an unstable environment and less work done by GAI recipients. Is there a way to get the desired outcomes you are looking for while we have a welfare system?

Justin writes:

GAI, considered as a replacement for the welfare state, simply won't work. There are too many goals that are inconsistent with each other. GAI needs to be affordable, bring everyone out of poverty, avoid discouraging work and be politically possible. Any attempt to reasonably balance these objectives will simply recreate a welfare state of targeted programs.

Jay writes:

@ Scott Sumner

Oh I agree, I just don't think Mr Justice's example was a particularly good one since there are ten things he could immediately change to improve his situation but chooses not to before even getting to a GAI.

Personally I think GAI as a welfare replacement is a good goal but like the flat tax will unlikely ever even come close to being considered since too many existing bureaucracies have funding tied to the myriad of existing welfare programs that they're unlikely to give up without fighting tooth and nail.

Jay writes:

@JJ

My guess is that you could start with the EITC and expand it into the GAI so that it would be structured to reward more for those who chose to work more instead of stay home, living off the check, and presumably be making babies.

Tom West writes:

Wait, are you saying that the solution of having the government give money directly to the populace is vastly *less* effective than having the government spending the money itself on the populace's behalf?

On this blog? :-)

Brett writes:

I'm skeptical this actually led to less employment, or the lack thereof would have led to more -unemployment is extremely high on native reservations with and without GAI.

I've grown more fond of a Job Guarantee/Job Subsidization program, but I still like a Basic Income Stipend idea (the key being "stipend", since a Negative Income Tax is a poverty trap).

This is due to the fact that people with a lot of free time tend to have more babies. And children that grow up in welfare households are much less likely to be employed as adults.

People on welfare don't have any more children on average than people who aren't, notwithstanding the anecdotes everyone supposedly has about the single mother on SNAP with 37 kids by 57 fathers.

Jim Rose writes:

Another effect of American Indian casinos is an increase in high school dropout rates because of a guaranteed job in the casino. This occurs even when the tribe offers college scholarships.

Tom West writes:

Another effect of American Indian casinos is an increase in high school dropout rates because of a guaranteed job in the casino.

Well, I'm certain Bryan would approve.

JJ writes:

Why couldn't Welfare be substituted for a Works program like WPA? Pay what you currently pay in Welfare(plus a small premium) and have them volunteer to do work like cleaning up blighted neighborhoods or whatever society needs.

Tom Davies writes:

So the good news is that living below the poverty line is not too bad?

JK Brown writes:

Hidden in that article is also a comment on government provided healthcare.

First we have this:

Native Americans have high rates of obesity and diabetes, which are often aggravated by a lack of good medical care.

Then in the conclusion they add this:
With free housing and health care, "a lot of people have figured out a way to use the system to survive," he says. "Why get a job if you don't need one?"

Free health care, most often via the Indian Health Service, but with conditions aggravated by lack of good health care...

TravisV writes:

Great find, thanks for sharing!

An igyt writes:

Is that "poverty rate" the usual one, based on income before taxes and transfers? If so, then the data compared says nothing about transmitting bad habits between generations, or anything nefarious like that -- it's just rational leisure choices. Like that of Mike Justice in the lead.

Serioulsy, why does anyone have beef with this guy's choices? He is staying with a son who would otherwise be raised almost alone by a recovering addict, and helping that mother recover, rather than spending half his day in a commute. I would bet that toddler has better odds in life because of it.

That's one point for the GAI. If the poverty data was consumption based, that's one point against.

Andrew_FL writes:

@Michael Byrnes-That line of reasoning won't work if they index it to inflation.

And in any case, now you've got a new source of output volatility.

brendan riske writes:

If you don't allow the market to work, people turn to political economy to get what they need. Everyone in the us wants a bailout now.its only fair.

Brett writes:

On topic, here is a response criticism of the article. Specifically read paragraphs #5 and downward:

We took a closer look at the law review article that The Economist relied on and were not impressed. It purportedly shows that poverty was more likely to increase in certain Pacific Northwest tribes that distributed part of their gambling revenues to members than in those that did not. But there were only seven tribes (out of a total of 17 that the article focused on) that did not distribute gaming revenues directly to members. The total reported decline in poverty among these seven tribes amounted to only 364 people. The study contained no controls for any of the many factors that affect poverty rates, nor did it take into account size differences in the tribes, differences in the size and structure of the per capita payments, or other relevant factors. In short, the study is absolutely useless in terms of providing meaningful evidence to support The Economist’s claim.

Even worse, The Economist failed to mention the existence of rigorous, peer-reviewed research contradicting the article’s thesis.

Hazel Meade writes:

@JJ,
I don't think there is any sort of welfare or GAI that won't disincentivize work, because work disincentivizes itself. Nobody likes working if they can help it.
The only reason many people work is because if they don't, they will starve and die. And many people will only do the minimal amount of work needed to avoid starvation and death, or maintain some bare minimum standard of living. So ANY amount of money you give them is going to disincentivize work since it gets them closer to the bare minimum beyond which they will stop working.

Hazel Meade writes:

People on welfare don't have any more children on average than people who aren't, notwithstanding the anecdotes everyone supposedly has about the single mother on SNAP with 37 kids by 57 fathers.

Not according to the 2006 census.

http://www.census.gov/prod/2008pubs/p20-558.pdf

For the nation, in 2006,10 years after passage of the Act,the birth rate for women 15 to 50years old receiving public assis-tance income in the last 12months was 155 births per 1,000women, about three times the ratefor women not receiving publicassistance (53 births per 1,000women).

Tom West writes:

I don't think there is any sort of welfare or GAI that won't disincentivize work, because work disincentivizes itself. Nobody likes working if they can help it.

As a generalization, I'd agree. But my personal observation (heavily middle-class) is that those who don't work, even if they don't have to, take a massive hit to their self-esteem.

Even those who have retired seem to suffer once they feel they're no longer necessary. Whether this is cultural or built into humans, I don't know.

So yes, we might not like to work, but we're miserable if we don't. (Again, by observation, unpaid work fulfills that need to be useful.)

Andrew_FL writes:
As a generalization, I'd agree. But my personal observation (heavily middle-class) is that those who don't work, even if they don't have to, take a massive hit to their self-esteem.

We can easily model this sort of feeling as a subjective utility function of labor

U(L) = SE(L) + I(L)

Where SE(L) is a function that starts at zero at zero work done, increases initially to a maximum at the amount of work a person would be willing to do for the self esteem alone, then diminishes as L increases and eventually goes negative, reflecting the dis-utility of labor beyond a certain point; At some point, it's simply not fulfilling to work for nothing any more, and instead it's just exhausting or otherwise unpleasant. I is the utility from income as a function of labor. Ordinarily, I is zero at zero Labor: with a GAI, it would be positive, and for some people, the maximum value of U they can achieve will be at L = 0. Some people. The people at themargin.

In other words, U(L) is not necessarily a strictly increasing function, but ordinarily it's zero at zero work. And therefore any amount of work, if any amount of work is available to be done, is preferable (ie has a higher subjective utility).

ChrisA writes:

Scott - isn't the case that almost everyone with the tendency to live off just the BI and not get a job is able to do that already given the current welfare system? This is more true in Europe if not in the US. And labor force rates in Europe are not significantly lower than the US (in the UK they are higher).

The downside of the current system is that there are many people who are caught in a welfare trap, who would like to work (perhaps only for a limited amount) but face very large effective tax rates if they do so, so they stay inactive. So yes, with a BI some people would idle, but perhaps more people would enter the workforce as a result.

Scott Sumner writes:

Michael, I agree with many of your points, especially the claim that a GAI might be better than our current system.

Of course the GAI could be "gamed" as well.

JJ, I think it's one of those "lesser of evils" issues. All proposals have their drawbacks.

Tom, No.

Brett, The finding that I found most interesting was the claim that the reservations that rely on the GAI approach have more poverty than those that rely more on getting people jobs.

JK, Good find.

An igyt, I have no "beef" with his choices.

Brett, Thanks for pointing that out.

Tom, Generally those on the left think a lack of work is extremely demoralizing if associated with the business cycle; particularly austerity, tight money, etc. But if the unemployment is caused by a GAI, their attitude shifts to "What's not to like?" Or at least that's the impression I get.

Hazel Meade writes:

@Andrew_FL:


I think things are a bit more complicated than that though. As I was saying, for some people additional I doesn't increase U beyond a certain threshold.

So it would me more like:
U(L) = { SE(L)+I(L) if I(L) SE(L)+ X else

Now supposing that for this individual SE is on the downward slope. And you add a certain amount onto I(L) with the GAI. If I(L) exceeds X now, suddenly this person can increase U(L) by working less.

And for SOME people, X is a pretty small number.


Hazel Meade writes:

Argh ...
U(L) = { SE(L) + I(L), if I(L) is less than X
SE(L) + X , else

Sorry the formula got messed up.

Andrew_FL writes:

@Hazel Meade-I had I(L) as the utility from income, as a function of labor, it's not income as a function of labor (that would have the wrong units, since satisfaction is dimensionless and subjective, where income is measure in money units).

I should have been clearer though, that I(L) is not necessarily a strictly increasing function of work done, either, even though money income may be.

And yes, I imagine many people can increase their subjective utility by working less than the greatest amount of labor which they can do. This is why people don't work themselves to death. Although, one could subsume, as I tried to do, all the dissatisfaction from labor into SE(L). My entire point was that the local maximum of U(L) in the range of L that is possible for a person to do (constrained by both their physical limitations, and by what people are willing to hire them to do) is not necessarily at the highest L possible. And that's more likely to be true, for people who can increase their money income by working less.

Mike Sax writes:

"Ever since the 1970s, I've favored wage subsidies for low wage jobs, and I've opposed the minimum wage and the guaranteed annual income (GAI.)

Many liberals actually have come to prefer a job guarantee

http://www.cfeps.org/pubs/wp-pdf/WP50-Tcherneva-Wray.pdf

Most conservatives in my experience though don't seem to prefer that. I seem to recall you being critical of the program even though it deals with the worry of discouraging work.

Comments for this entry have been closed
Return to top