Universal Basic Income: a thoroughly researched, evidence-based policy

Image: President Richard Nixon bowling (source)


Poverty--by definition--is a person or family with not enough money. If poverty is simply not having enough money, what if we just gave people free money? Wouldn't that solve the problem of poverty?

Today, this proposition sounds completely absurd. No politician or commentator in the mainstream could even mention it and be taken seriously. This is the kind of proposal that could end a career. Yet, decades ago, this idea was so mainstream that the likes of Richard Nixon and Milton Freeman were in support of it (this is a useful reminder of just how conservative our country has become). Yes, as Nixon was escalating in Vietnam, he represented the political center by advocating a universal basic income.

The idea has many names, among them Universal Basic Income (UBI), Guaranteed Annual Income (GAI), or (as Nixon would have known it) the negative income tax. It also takes several different forms: some proposals would pay everyone the difference between their earnings and a poverty line--for example, if the poverty line was $20,000 in a year and your family only made $16,000, the federal government would write your family a check for $4,000. Others aim to keep things administratively very simple and would direct deposit a set amount of money in the bank accounts of every single citizen* (for most people, these benefits would be offset by increased Social Security taxes to pay for the program). In whatever form, a UBI is a guarantee that you and your family will never be impoverished.

A hallmark of all UBI proposals is unconditional cash transfers--that is, recipients are not expected to do anything in order to receive cash benefits. Again, this makes the program administratively very simple, but it also grants people security. That security has its own rewards (which I will elaborate on in a future post). (Even conditional cash transfers in developing countries are often considered utopian nonsense--yet they work).

It's crazy to think that a UBI would improve health or educational outcomes. And it's equally crazy to think that a UBI would not cause people to stop working. Yet that is exactly what research on UBI's has found. Decades ago, when UBI was a serious policy proposal, extensive research was done. Very well designed studies (randomized controlled trials) found downright incredible results. From a review:

The first experiment was conducted on an urban population in New Jersey and Pennsylvania between 1968 and 1972. A second experiment was conducted in Gary, Indiana to examine the effect of a GAI on single parents. A third experiment was conducted in North Carolina and Iowa to look at the effects on rural populations. The final experiment was the Seattle-Denver Income Maintenance Experiment (SIME-DIME) which had access to a much larger experimental population...

In North Carolina, children in experimental families showed positive results on elementary school test scores. In New Jersey, data on test scores was not collected, but a positive effect on school continuation rates was found. In SIME-DIME there were positive effects on adult continuing education (Levine et al. 2005: 100). These results are all the more remarkable when juxtaposed to the academic literature that shows it is very difficult to affect test scores, dropout rates or educational decisions by direct intervention.

Inconsistent attempts were made to collect health data, specifically on issues such as low birth weight which can be associated with significant deficits in later life. The Gary, Indiana study found positive effects on birth weight in the most at-risk groups (Levine et al. 2005: 100).

And here's what those studies found on the question of whether work was affected:

The experiments generally found a 13% reduction in work effort from the family as a whole, with one-third of the response coming from the primary earner, one-third from the secondary earner and the final third coming from additional earners in the family (Levine et al. 2005: 99). Because the primary earner typically worked many more hours than the secondary and tertiary earners, this implied a relatively small reduction in work effort by primary earners. Female spouses reduced their hours and re-entered the workforce less quickly after a break. The general result that secondary earners tend to take some part of the increased family income in the form of more time for household production, particularly staying home with newborns, was found in all the experiments. Tertiary earners, largely adolescent males, reduced their hours of work dramatically but the largest decreases occurred because they began to enter the workforce later. Taking a first job at an older age suggests that some of these adolescent males might be spending more years in school. The biggest effects, that is, could be spun as either an economic cost in the form of work disincentives, or an economic benefit in the form of human capital accumulation.

In sum, all studies found that a typical family involved a primary earner (a parent--usually a father--with full time job), a secondary earner (a parent--usually a mother--with a part time job), and one or more tertiary earners (children in high school with after school jobs). In aggregate, primary earners cut back their hours only slightly; secondary and tertiary earners cut back their hours in a small but significant way, but usually did so for reasons we would likely approve of (such as a substitute for maternity leave or to return to school for secondary earners; cutting back on hours to spend more time studying--and thus graduating from high school--for tertiary earners).

[see update below]

UBI was likely killed by a mistake a SIME-DIME researcher made (emphasis added):

The most damning result, however, came in the form of somewhat questionable family dissolution rates in the SIME-DIME experiment. These results seemed to imply that black experimental families had a divorce rate 57% greater than the controls, and [sic] while experimental families had a divorce rate 53% greater than the controls...Further analysis of the data, published in 1990, rejected these findings as a statistical error, and no other experiment found any effect on marital stability (Cain 1990).

In addition to these studies, the Canadian government financed a UBI study of their own. They called their proposal "Mincome". The Mincome study actually had several experiments running simultaneously in different locations. Most experiments were designed like the American studies reviewed above, and found similar results:

[U]ltimately the findings were very similar to US findings: secondary and tertiary wage-earners tended to have a moderate labor market response, while primary earners showed little change in the number of hours worked[.]

It found that in aggregate, the number of hours worked fell "about one per cent for men, three per cent for wives, and five per cent for unmarried women." Again, the decision to decrease work hours generally occurred for reasons we would likely approve of (ie, as a substitute for parental leave, to go back to school, or to stay in high school to graduation).

One study of Mincome was unique: the small city of Dauphin, Manitoba (approximately 10,000 families), was given a truly universal basic income. This made the study groundbreaking for two reasons. First, most experiments were narrowly designed to study the effect of a unconditional cash transfers on work incentives; thus, the elderly and disabled were ineligible to participate in all other American and Canadian UBI studies. Second, most studies were designed to have control groups in the same city, so all test sites had a mix of participants with and without grants. In Dauphin, however, no one was excluded from the GAI. Every single person in the city received grants.

Major takeaways of the Dauphin Mincome experiment--in comparison to control group (a nearby town):

• Significant reduction in hospital admissions, particularly mental health- and accident-related admissions
• Decline in physician contacts for mental health diagnoses
• Decline in teenagers dropping out of high school
• People did not have more babies
• No increase in family dissolution

In their abstract, the authors claim that they were able to study the affect on birth outcomes, but they later explain that they were not able to do so:

One of the strongest claims made by individuals who advocate a GAI is that birth outcomes may be improved: improved nutrition and access to prenatal care may lead to healthier newborns. The only data we were able to use to determine birth outcomes were perinatal deaths – extremely rare events – and birthweight. The detailed birth records of recent years, with 1- and 5-minute APGAR scores and other data were not available in the mid-1970s.

In other words, the only useful, available data on birth outcomes was perinatal death, but this is too rare an event for their data to be meaningful. Since the city had just 10,000 people in it, the number of perinatal deaths was too small to draw any conclusions. Unfortunately, no other useful data were available to researchers regarding birth outcomes. When this problem was studied more thoroughly in Gary, Indiana, it was found that there was no effect on birth outcomes for most mothers, but that there was a significant, positive effect for at-risk mothers.

More studies
More recent studies have looked into the effects of UBI, usually finding (as previous studies have) that our common understanding of poverty reverses causality. Poor educational outcomes, mental health problems, and criminal activity don't cause poverty; poverty causes these outcomes (we're talking about risk here--poverty doesn't guarantee poor outcomes but makes it more likely). Example:

[I]n 1996, the Eastern Band of Cherokee Indians in North Carolina’s Great Smoky Mountains opened a casino, Jane Costello, an epidemiologist at Duke University Medical School, saw an opportunity. The tribe elected to distribute a proportion of the profits equally among its 8,000 members...By 2001, when casino profits amounted to $6,000 per person yearly, the number of Cherokee living below the poverty line had declined by half.

The poorest children tended to have the greatest risk of psychiatric disorders, including emotional and behavioral problems. But just four years after the supplements began, Professor Costello observed marked improvements among those who moved out of poverty. The frequency of behavioral problems declined by 40 percent, nearly reaching the risk of children who had never been poor. Already well-off Cherokee children, on the other hand, showed no improvement.

Minor crimes committed by Cherokee youth declined. On-time high school graduation rates improved. And by 2006, when the supplements had grown to about $9,000 yearly per member, Professor Costello could make another observation: The earlier the supplements arrived in a child’s life, the better that child’s mental health in early adulthood.

Another truly universal UBI was studied in the rural village of Otjivero-Omitara, Namibia, with similarly positive results using extremely small grants:

The BIG [Basic Income Grant] Coalition raised money which allows them to give a BIG of 100 Namibian Dollars to each individual which was registered in July 2007 as living in the Otjivero-Omitara area, about 100 kilometres east of Windhoek (pensioners were excluded as they get an unconditional state pension). The amount is small, since the food poverty line stands at 152 Namibian dollars per capita, whereas the poverty line counting “the severely poor” stands at 220, and the official poor are all those living on less than 316 Namibian dollars per month.

As the study of the effects of the BIG after one year clearly demonstrate, the effects are strikingly positive. The percentage of those falling below the food poverty line has dropped from 76% to 37%. The percentage of those being able to get a job or become successfully self-employed has increased from 44 to 55%, and the amount of non-BIG income per capita rose from N$ 118 to N$ 152 (indicating a virtuous economic growth cycle). The number of underweight children has dropped from 42 to 10%. School attendance has gone up, and teachers report that the children are better able to concentrate. The health clinic receives many more patients (for illnesses that would otherwise not have been treated). Average household debt fell from N$ 1,215 to N$ 772. Crime rates fell by 42%, and there is no evidence that alcohol-abuse (which is a serious problem in many poor areas) has worsened. (Further details are in the report, together with interviews documenting the experiences of the people who have been given the BIG).

Incredible. Read that entire blockquote again. Note that non-UBI income increased. This clearly indicates that rather than being less productive and becoming dependent on their cash grant, people actually became more productive. Paradoxically, when people don't have to worry about the basics, they can be more productive. Fear of hunger and destitution doesn't make you more productive for fear of the consequences; it makes you less productive as your mental resources are taxed by your fears and you go into survival mode (much more on that in a future post). Despite the extremely paltry size of the grants, more businesses were created, possibly because the additional security encouraged people to take risks and start a business. Clearly, a UBI is an extremely cheap and effective way to enact positive change in the lives of the poor in underdeveloped countries.

(UPDATE 07/01/2014: Dylan Matthews has much more on research that found that cash aid to the poor--even when those poor are actually drug addicts--does not get spent on drugs, alcohol, or other vices)

GiveDirectly is a charity that provides unconditional cash transfers in Kenya, though these are not universal. A randomized controlled trail of this program found--among other promising results--that the transfers decreased hunger, did not result in increases in spending on alcohol or tobacco, increased investment in and revenue from livestock and small business, and reduced participants' levels of stress hormones. (This American Life recently did an entertaining segment on GiveDirectly. As usual, A+ for the title: Money for Nothing and Your Cows for Free.)

None of this should be surprising. Nearly all research we have on work and incentives argues that humans have an innate need to contribute to society and be productive; incentive payments are usually harmful to worker performance. People aren't poor because of their vices--that's why simply giving them money doesn't result in them spending more money on vices.

More Canadians than not support a UBI (h/t). The Swiss will have a referendum on implementing a UBI in 2014. Matt Breunig created an interesting calculator using census data to estimate the cost and effect on poverty of different UBI benefits. Few public policy ideas have such a large evidence base behind them and the potential to help so many people.

More strengths of UBI will be discussed in later posts.

Related:
The one graph that proves that welfare is not a work disincentive


Update 07/23/2014
Dylan Matthews addresses some extremely misguided interpretations of the results of the RTC's done in the 1960's and 70's--namely, the suggestion that the studies imply catastrophic effects on the economy. He also makes a salient point about the cost of the program; even assuming large effects on workforce participation, an investment of just 1.5% of total United States GDP would eliminate official poverty.

Update 10/03/2014
Dylan Matthews continues his excellent work aggregating research on UBI and unconditional cash transfers. Of a study of Syrian refugees, Matthews writes (emphasis in original):

1. Those getting the payments spent more on heating fuel and clothes, and were "significantly more likely to own an oven and heater."
2. They were less likely to skip meals or have adults eat less to help children.
3. They were less likely to engage in child labor or physically dangerous work, or to sell productive assets.
4. Children in recipient families were more likely to go to school.
5. Families getting cash were less likely to report domestic disputes.
6. Those getting the cash worked a tiny bit less but there's barely any work to go around: recipient adults worked, on average, 2.7 hours in the past four weeks, compared to 3.1 for those not getting the cash.
7. Recipients spent no more on candy or beverages and significantly less on tobacco; once again, there's no evidence that poor people blow cash transfers on booze and cigarettes.
8. Survey responses suggest the cash "did not lead to increased crime or jealousy."

(Much more on the methodology and positive effects on the economy)

Update 01/18/2015
A randomized, controlled trial in rural India (all residents--adults and children--in 8 villages received a UBI, with 12 villages used as controls) found extremely positive results:

Positive results were found in terms of nutrition, health, education, housing and infrastructure, and economic activity. The researchers found that the cash transfer group spent significantly more on eggs, meat, and fish than the control group. Researchers found a positive impact on health and access to medical treatment. The most visible impact of the study was on educational attainment. Researchers found increased spending on school-related items such as school uniforms, school fees, shoes for school, books, school supplies, and private tuition. School attendance in the cash transfer villages shot up, three times the level of the control villages. Performance in school improved significantly relative to control villages. There was increased investment in housing, such as the installation of in-door plumbing.

*A major worldwide organization, the Basic Income Earth Network, advocates for a worldwide UBI. They calculate that

Paradoxically, therefore, giving to all is not more expensive but cheaper than giving only to the poor.