Karl Widerquist, A Critical Analysis of Basic Income Experiments

Karl Widerquist, A Critical Analysis of Basic Income Experiments for Researchers, Policymakers, and Citizens, Palgrave Macmillan, 2018, xi + 167 pp, 3 030 03848 9, hbk, £49.99

One of the causes of the current lively debate on Citizen’s Basic Income, at least in the UK, was a piece of fake news: the news, about three years ago, that Finland was about to pay a Citizen’s Basic Income to all of its citizens. The result was a discussion of Citizen’s Basic Income on BBC2’s Newsnight, radio interviews, and a Radio 4 Money Box Live programme. The source of the fake news might have been an article in the New Statesman that simply got it wrong. What had actually been promised was a pilot project. That began in January 2017 and finished in December 2018, initial results have been published, and we await further results. This was of course nothing like as significant as a Citizen’s Basic Income for every individual in Finland, but paying an unconditional income for two years for two thousand randomly selected unemployed individuals was still an important event. Pilot projects matter. It is therefore a pleasure to see a book entirely about Citizen’s Basic Income experiments from one of the acknowledged authorities on Citizen’s Basic Income.

Widerquist’s book’s introductory chapter asks that researchers should plan their experiments in relation to the questions that they wish to answer. Chapter 2 distinguishes between Citizen’s Basic Income and Negative Income Tax, and this chapter and chapter 5 conclude that the latter is easier to experiment with in more developed countries. The extent to which the results of a Negative Income Tax experiment can tell us what would happen if a Citizen’s Basic Income were to be implemented is debatable.

In chapter 3 Widerquist discusses terminology and testing methods. The book explicitly avoids ‘pilot project’ terminology, because that might imply that the policy being tested will be implemented. Instead, ‘experiment’ terminology is employed with a broad definition. A much tighter definition is employed for ‘Randomized Controlled Trial’ (RCT), with Widerquist declining to use it for saturation studies, in which a whole community receives Citizen’s Basic Incomes, because comparing a single community subject to a new policy with one not subject to it does not constitute the random selection of the statistically significant sample of communities that would be required for an RCT. The Finland selection of two thousand individuals from across the country did constitute an RCT, whereas although the Namibian experiment was not one, and there was not even a control village, the results obtained can still be regarded as significant. Widerquist’s view is that both RCTs and saturation studies are useful for different reasons, and that both need to be carried out: a good example of the balance to be found in much of this book.

Chapter 4 describes some of the common difficulties encountered in any social experiment (such as the Hawthorn effect, in which an experiment can end up measuring the effect of researcher observation on a community rather than the effect of the policy ostensibly being tested). Chapter 6 reviews the US and Canadian Negative Income Tax experiments of the 1970s, and shows how contemporary political and public misunderstanding of the results meant that the positive outcomes were only appreciated more than thirty years later, as the beginning of the all too brief chapter 7 shows. The rest of chapter 7 reviews the conduct and results of the important Namibian and India saturation studies. Chapter 8 discusses some current experiments, and warns that researchers should be prepared for negative employment effects – so it is of interest that the first results of the Finland experiment, released after the publication of this book, show no employment effect, which might suggest that the employment-reducing effect of the secure income was balanced by the employment-enhancing effect of lower marginal deduction rates. We await further results from the Finland experiment with interest.

Chapter 9 is a useful discussion of the reasons why Citizen’s Basic Income experiments might be held. The last reason listed might be the most significant: politicians agree to an experiment because they feel a need to respond to a growing demand for a Citizen’s Basic Income, don’t wish to risk implementing one, and know that holding a small experiment will successfully put off further discussion of implementation of a nationwide Citizen’s Basic Income until the experiment has been evaluated. Chapter 10 reiterates the vulnerability of experiments to misunderstanding of the results, and develops a point already made in chapter 1, that researchers should keep the questions to be answered in mind and should therefore undertake a constant dialectic between the questions to be answered and the design of the experiment. Precisely which questions are being addressed is a crucial question, and it will always need to be made clear that experiments can only answer a limited subset of the questions that various interested parties might be asking.

Chapter 11 argues that public debate about Citizen’s Basic Income is mainly about ethical concerns: an issue to which an experiment can contribute little – although perhaps more might have been said about the connection between experimental employment effects and ethical issues related to the presumed negative employment effects of Citizen’s Basic Income. Chapter 12 again emphasises the importance of being clear about the questions that an experiment might be able to answer, and about how research results will respond to those questions.

Chapters 13 to 16 list claims for and against Citizen’s Basic Income, and finds that experiments are not very good at responding to them; and chapter 17 encourages researchers to undertake a feasible experiment as close as possible to the one that they wish to carry out if the experiment that they might wish to carry out is not feasible for one reason or another. Chapter 18 addresses the question as to whether experiments should be held, notes that they are being held, and suggests that the best possible use should be made of those being carried out. The final chapter suggests that experimental results should be published in such a way as to respond the questions that interested parties are asking.

A clarification: Widerquist accuses microsimulation of being a ‘highly imperfect method’ (p. 123) in relation to employment market behaviour. Yes, it is, because if the microsimulation model is used to calculate marginal deduction rates, and those are then employed to estimate behavioural change, then it is in that final step that presuppositions enter the evaluation, and those can be flawed. For other purposes microsimulation is the best method available, and in fact functions like a practical experiment – which might have been said. A microsimulation programme has coded into it a country’s tax and benefits regulations, and it has passed through it financial data on a massive and statistically significant sample of the population (in the UK, the sample is 0.1% of the entire population). More accurately than any other method, microsimulation can tell us how much a Citizen’s Basic Income scheme would cost, how it would affect poverty and inequality indices, how many household income gains and losses it would cause, and where in the income range those would occur, how many households would no longer find themselves on which benefits, and so on. If those are the questions that need to be answered, then microsimulation is the best experiment available.

A minor quibble: Whether the unconditional income implemented in Iran by accident, or such existing unconditional incomes as the UK’s Child Benefit, should be regarded as Citizen’s Basic Income experiments is an interesting question, but they should at least have been discussed.

Widerquist’s book will not be the end of the sometimes vociferous arguments about the usefulness of experiments and about how they should be carried out, but it will be essential reading for anyone planning an experiment. It is an important contribution to the Citizen’s Basic Income literature.