Keywords

The belief that a UBI experiment can provide a definitive answer to the question of whether to introduce UBI rests on three false presumptions: (1) People disagree about UBI primarily because they disagree about what its effects might be. (2) These disagreements about effects stem from a lack of available evidence. (3) An experiment will provide that missing evidence.

In each case, nearly the opposite is true: (1) Although some important disagreements about UBI’s effects exists, the disagreement is more of an ethical debate about the desirability of its effects than an empirical debate about what those effects are. (2) Disagreements about what UBI’s effects are don’t stem primarily from a lack of available evidence. Substantial evidence already exists. Some of it is widely known; some isn’t. (3) Experiments cannot provide the most important missing evidence. They will only add a small amount to the existing body of evidence and leave many important empirical questions about UBI unanswered.

Therefore, this chapter explores the difference between the questions that need to be resolved to make a decision for or against UBI and the answers UBI experiments can provide. The difference is bound to disappoint some readers, and some might react by saying why bother (see Chap. 18), but it’s necessary to understand what an experiment’s limits are if we’re going to get the most out of it.

Experiments are empirical studies. They can provide evidence to help answer empirical questions like, what does this do? But they cannot provide the answer to ethical and subjective questions, such as do we want what this does? Experiments cannot resolve the basic disagreement, which is more about the second question than the first. The focus on ethics is not because people don’t care about evidence, but as argued below, because UBI’s likely effects are well-enough understood and the moral desirability of those effects is controversial enough to make the ethical part of the argument pivotal.

For example, UBI supporters tend to believe either (1) that it is good for everyone to be free from the threat of poverty, including nonwealthy people who might refuse to take jobs, or (2) that the possibility of nonwealthy people refusing to take jobs is not bad enough to compel sacrificing other goals that UBI achieves. Opponents tend to believe it is wrong for anyone (who isn’t independently wealthy) to get anything without taking a job. These positions differ on basic ethical premises—as do positions in many similar disagreements over UBI. No empirical study of the practical effects of a UBI will determine whether these two incompatible ethical beliefs are right or wrong.

Although there are many nonethical reasons to support or oppose UBI, this ethical divide exists in the background of most discussions over UBI’s ability to achieve its goals. People who haven’t made up their minds on UBI often bring up concerns that are closely related to this and similar ethical disputes.

This aspect of the UBI discussion makes it very different from the climate change discussion. One reason the denial of evidence plays such a large role in the climate change debate is that if climate change is happening, it seems obvious we should do something about it. Therefore, those who don’t want to do anything about it feel they have to get people to believe it is not happening. By contrast, it is entirely possible for two people to agree about all the effects of UBI and disagree about whether it is a good policy. People on one side of the issue or the other do not necessarily have to deny any evidence to make their case for or against the policy.

Empirical research can find evidence that is useful to people discussing ethical issues. For example, if research was to find out that people with UBI tend to make as good or better contributions to society as people do without UBI, at least some people who are leery about allowing nonwealthy people to live without taking a job would probably become more open to UBI. But not all people who oppose UBI for this kind of reason will be swayed. Some might believe nonwealthy people need to work more than they are working now. Others might oppose UBI because they oppose even the possibility that a nonwealthy person might refuse to participate in the labor market.

Similarly, if empirical research found that a given level of UBI caused a decrease in employment so large that it threatened UBI’s sustainability, any UBI supporters who aren’t extremely short-sighted would drop their support for UBI or at least for that level of UBI if they were unable to suggest a policy to counteract that unsustainability.

Yet, experiments in wealthy countries are unlikely to show either result. Past evidence strongly indicates that low-wage workers in wealthy countries will spend less time in employment, but not so much less that UBI will become unsustainable. If experiments are consistent with a decline in labor hours in that range, supporters are likely to say UBI passed the test and opponents are likely to say that it failed. People whose opinions are in the middle might be swayed by where in that range the estimate falls, but a subtle finding like this isn’t likely to be a huge deviation from what we can already estimate from existing evidence. And it’s possible that responses of people in the middle will be affected less by the small amount of additional evidence than by who wins the spin wars that are likely to follow the release of experimental findings.

Closely related to the issue of whether empirical findings can resolve the ethical debate over UBI is the problem of separating empirical from ethical claims about UBI. Almost any social policy study has to deal with the problem that it is not easy to evaluate whether the policy “works” without making ethical judgments about how to evaluate performance. We can say empirically whether UBI meets criterion X, but we have to make an ethical judgment to say how important X is as a criterion (see Chap. 12).

An enormous amount of evidence about UBI’s effects already exists. Thousands of articles and books on various aspects of its effects have been written and seven large-scale trials conducted worldwide between 1968 and 2013. In addition, studies of full-fledged policies of varying degrees of similarity to UBI, such as the Alaska Dividend, conditional cash transfers , citizens’ pensions, tax credits, and many others, provide information that can be used to estimate UBI’s effects.Footnote 1

My impression—after studying UBI for more than 20 years—is that the better one grasps existing evidence, the more likely one’s decision comes down to ethical issues. I can say that the right UBI scheme will be sustainable and will do things people with an ethical position similar to mine want it to do, but it will also do things that people with different ethical positions do not want it to do.

Although many reasonable people are in the middle and might well be swayed by new evidence, many people in the middle aren’t familiar with the existing evidence, and it is uncertain that a new experiment will provide the most important piece of missing evidence they have been looking for.

Existing evidence is not assembled in any one spot nor is most of it easily accessible to nonspecialists. The most accessible summaries of existing evidence are in books written by supporters such as Annie Miller, Guy Standing, Malcolm Torry, Philippe Van Parijs, Yannick Vanderborght, and others.Footnote 2 Of course, books by supporters might be subject to confirmation bias.

Despite the enormous amount of evidence available in the relevant social science literature and the availability of good summaries, a substantial part of the current discussion of UBI among citizens and policymakers still goes on in ignorance of existing evidence. In fact, a lot of clearly false claims easily contradicted by evidence are regularly repeated in the debate. For example, many people continue to claim that a poverty-level UBI would cost 15–20% of gross domestic product (GDP), when the actual amount is estimated to be about one-sixth of that figure, less than 3% of GDP.Footnote 3 Future discussions might go on in ignorance of most of the findings of the current round of experiments.

Important gaps in existing evidence do remain. Experiments can help fill in some of those gaps, but as following chapters discuss, experiments are only capable of testing a small subset of what we really want to know about UBI. And many of the biggest and most important gaps in the existing evidence are not things that UBI experiments are capable of addressing. Neither these gaps nor the potential for UBI experiments to fill them in are well-understood by nonspecialists, including some of the reporters currently writing about the experiments.

The decision to conduct a UBI experiment should be made with full knowledge of all these limitations. If we want a UBI experiment, we need to accept not only that it is incapable of settling the major ethical divides between supporters and opponents, and that it is highly unlikely to prove either of their positions untenable, but also that it is unlikely to provide a large enough addition to existing evidence to give a compelling reason for massive numbers of people in the middle to shift their opinions significantly. Like most social science research projects, UBI experiments will make an incremental contribution to existing evidence. If the results are well-communicated, their best realistic hope is to enlighten the discussion among people on all sides of the current discussion by increasing both the evidence available to them and their understanding of it. This is a good reason to do an experiment, but it is far short of the definitive test people want and some seem to be expecting.

I suspect that some specialists will mistakenly believe that everything this chapter says is too obvious to mention. That mistake is a central reason this book is necessary. Citizens and policymakers have to be free of false hopes if their decision to conduct a UBI experiments is to be based on what an experiment can actually do.