This post by Edward Freeland, Associate Director of the Survey Research Center at Princeton University, presents results from a recent survey of public opinion on a Universal Basic Income. We find that support and opposition are roughly equivalent, but responses can be pushed in one direction or another by the order in which questions are presented.
The concept of a Universal Basic Income (UBI) is turning up more often these days in debates over income inequality, poverty and the impact of automation on the labor force. It’s an old idea that traces back to Sir Thomas More’s Utopia and appears later in the writings of social philosophers such as Condorcet, Thomas Paine, Charles Fourier, and John Stuart Mills. In the 1960s, economist Milton Friedman proposed a “negative income tax” that would replace America’s patchwork system of social welfare programs with unconditional cash payments that would be enough to cover the basic costs of food and shelter for low income households. Friedman’s idea was the basis for the Family Assistance Program that was proposed (and then defeated) during the Nixon Administration. Recently the idea has drawn renewed interest because of skepticism about the effectiveness of aid programs in developing nations and concerns about workers who will be displaced in the near future by robots and artificial intelligence. Several countries around the world and several cities in the US have announced their interest in UBI, and some have launched experiments designed to test its feasibility.
For the most part, public opinion on the subject of UBI has been divided. A 2017 poll by Gallup found a near even split among American adults between those who support UBI (48%) and those who do not (52%). A Zogby Strategies poll done around the same time found 40% in favor of UBI, 35% opposed, and 25% undecided. A 2017 international survey by IPSOS found a more even split among Americans, with 38% in favor, 38% opposed and 24% undecided. Among other nations in the poll, support for UBI was found to be lowest in countries such as France (29%) and Spain (31%) and highest in Poland (60%) and Germany (52%). Switzerland is the only country that has had a national referendum on this issue. In 2016, voters there rejected a proposal to give each person a monthly cash allowance of 2,500 Swiss francs (about $2,000 US dollars when you factor in the cost of living) by a margin of nearly 3 to 1.
We added four questions to the January/February 2019 Understanding America Study omnibus survey to get an update on opinions of UBI from a representative sample of American adults (UAS 167). We had four objectives:
- Gauge current public support for government sponsored UBI in the US
- Measure support for UBI programs sponsored by nonprofit organizations for people living in the poorest parts of Africa
- Measure how much people agree on society’s obligation to provide for the basic needs of all children regardless of where they are born
- Test the effect of asking about government sponsored UBI in the US conditional on having first thought about nonprofit sponsored UBI in the poorest parts of Africa
Here are the questions (the order of questions 1 and 2 was randomized):
1. Some business executives have proposed that the U.S. government should provide every American with a Universal Basic Income, that is a minimum amount of money each month so they can afford a basic level of shelter, healthcare and food. Do you support or oppose this idea?
- No Opinion
2. Some charities and international aid organizations have proposed using their funds to provide people living in the poorest parts of Africa with a Universal Basic Income, that is, a minimum amount of money each month so they can afford a basic level of shelter, healthcare and food. Do you support or oppose this idea?
- No Opinion
3. To what extent do you agree or disagree with the following statement:
All children, no matter where they are born, should be provided the opportunity to go to school, develop their talents, live safely and receive medical care when they need it.
- Agree strongly
- Agree somewhat
- Neither agree nor disagree
- Disagree somewhat
- Disagree strongly
In terms of public support for government sponsored UBI in the US, our results are in line with those found in previous polls: 37% support it, 40% are opposed and 23% have no opinion. Women are more likely to be UBI supporters than men (40% support vs. 35% support), people with bachelor’s degrees are more likely to be opposed to UBI than those with less education (46% opposed vs. 37% opposed); non-Hispanic whites are also less likely to support UBI than everyone else in the sample (31% support vs. 52% support). People below the median age for the sample (53 years) are more likely to support UBI than those who are older (41% support vs. 33% support). People with household incomes below the median are also more likely to be domestic UBI supporters.
On the question of non-profit funded UBI for people living in the poorest parts of Africa, support among adults in the US is somewhat stronger, although a higher proportion have no opinion: 41% support it, 25% are opposed and 34% have no opinion. When we look at the demographic differences explored above in question 1, we find similar patterns of support and opposition, although there is a big drop (23 percentage points) in opposition among the more highly educated, with 9% of people with bachelor’s degrees moving from opposing UBI in the US to supporting UBI in Africa and 14% moving from opposing UBI in the US to no opinion on UBI in Africa.
As noted earlier, we randomized the order of appearance for questions 1 and 2. Our null hypothesis was that the randomization would make no difference in whether one supports or opposes either type of UBI. However, we find that responses to both questions are indeed impacted by their order of appearance.
When the question about government-sponsored UBI in the US appears first, support for nonprofit-sponsored UBI in Africa decreases by seven percentage points (from 44% to 37%). This shift is consistent across all race/ethnic groups in the sample and appears to be unrelated to any differences in education, age or sex. Irrespective of demographic differences, being asked first about government-sponsored UBI in the US appears to make one less sympathetic to the idea of nonprofit-sponsored UBI for poor Africans.
Why this outcome? It’s possible that respondents misunderstood (or ignored) the difference in the source of funding for the two different forms of UBI. Perhaps government-sponsored UBI in the US was perceived as a domestic aid program and nonprofit-sponsored UBI in Africa was perceived as a foreign aid program. We know from other polls (e.g., the GSS) that while neither form of assistance is popular, spending on foreign aid is much less popular than spending on domestic welfare programs. So, it’s possible that the difference in geography loomed larger in the minds of respondents than the difference in funding sources.
But what happens when we ask first about support for government-sponsored UBI in the US? If we ask respondents first about this type of UBI, 33% say they support it. But when we ask first about nonprofit-sponsored UBI in Africa, support for government-sponsored UBI in the US rises to 42%, an increase of 9 percentage points. Thus, it appears that being primed to think about nonprofit sponsored aid in Africa makes one more inclined to support government sponsored UBI in the US. Here again, the difference in geography may have been more cognitively salient to respondents than the difference in the funding sources.
These two ordering effects (the negative effect for Africa when we ask first about UBI in the US and the positive effect for the US when we ask about UBI in Africa first) hold up even in multivariate logistic regressions of support for the two forms of UBI on a set of demographic factors where we include the order of appearance as a predictor variable.
As for our third question, the one about whether all children, regardless of where they are born, should have their basic needs met, 66 percent of respondents strongly agreed and another 20 percent agreed somewhat. Only eight percent said they neither agreed nor disagreed. Here we find a significant decrease in the percent of respondents offering a “don’t know” response. In addition, responses to this question are not significantly impacted by the order of appearance of the first two questions.
In sum, we find that our results are very much in line with the results of previous polls and surveys: support for and opposition to Universal Basic Income are roughly equivalent with a significant percentage of the sample offering a “don’t know” response. If, in the near future, the issue of UBI starts to gain political traction, public opinion researchers should be mindful of the apparent ease with which responses can be pushed in one direction or another by the order in which questions are presented. While support for UBI is likely to remain counterbalanced by opposition, advocates can take some solace in the strong support found among most Americans for the general principle that all kids, no matter where they are born, should have their basic needs met by the larger society.
NOTE 1: The survey questions were designed by Alan Krueger and Edward Freeland, with help from Jill Darling. Alan Krueger passed away in March 2019.
NOTE 2: Results described are based on weighted tabulations of responses from 3,824 respondents. Unless otherwise noted, differences in percentage outcomes are statistically significant at the p≤.05 level.