Some thoughts on basic income ‘experiments’

Some thoughts on basic income ‘experiments’

Michael A. Lewis

I recently read Kate McFarland’s very informative overview of several basic income “experiments.” The quotes are around that last word in my previous sentence because, as McFarland notes, not all these projects are truly experiments, at least not if the word “experiment” is being used the way it is in the social and biomedical sciences. As we use this term in the social sciences, an experiment is a study with the following features:

  • Study participants or a cluster of them are randomly assigned to at least two groups
  • At least one of the groups is a treatment group, while at least one is a control group
  • The treatment group receives the intervention of interest, while the control group does not receive intervention.

Feature one above is key.

What random assignment does is make it very likely that the treatment and control groups will be balanced. “Balanced” roughly means that the distribution of variables related to both the intervention and outcome of interest are the same across treatment and control groups. So if after the data are analyzed we find a difference in the outcomes between treatment and control groups, we can attribute such a difference to the intervention of interest.

The random assignment feature is why Eight’s study in Uganda, as McFarland points out, has limited “usefulness as an experiment.” I think it is fair to say in fact, that social scientists would not consider what Eight is doing an experiment at all. I am not saying that Eight’s study has no usefulness whatsoever. It may be useful when it comes to keeping BI “in the spotlight” and, thereby, help to maintain attention on this movement. For those of us who, at least in principle, like the idea of a basic income, this is a good thing. But we should be careful when it comes to considering what we can learn from the Uganda “experiment.”

The study in Uganda is usually called a pre-test/post-test study. In such studies, measures are taken before an intervention of interest (the pre-test part), after the intervention is implemented (the post-test part), and then these “before and after” measures are compared to one another. If certain changes are observed, these may be attributed to the intervention in question. The problem with such studies is that we do not know what would have happened to the group which received the intervention had it not received it. Maybe the observed changes in the relevant measures would have occurred even if there had been no intervention. The reason we want control groups in experiments is to allow researchers to estimate what would have happened to the group that received the intervention had it not received it. Without a control group, the Uganda study simply may not tell us much about the effects of the cash grants they are testing.

The third feature above has to do with the intervention of interest. This is very pertinent to the experiments McFarland wrote about, as well as BI experiments in general. Following BIEN, McFarland defines BI as “a periodic cash payment unconditionally delivered to all on an individual basis, without means-test or work requirement.” As I read her piece, I thought she was interpreting this definition to mean that if a policy provides a cash payment, exactly as spelled out in the definition, but also decreases the payment if a recipient obtains an income from selling their labor, then such a policy wouldn’t be a basic income. Alaska has no income tax, but it does have the Permanent Fund Dividend. Since it gives folks the dividend but does not tax any of it back in the form of an income/earnings tax, its grant would be an example of a basic income. But if the U.S. or any other nation, granted people money unconditionally, periodically, on an individual basis, and without a means test but also taxed all sources of income, including earnings, then that country would not have a basic income. This may seem like a mere semantic point, having nothing to do with BI experiments. But I think it is incredibly relevant.

McFarland makes it clear that some places are assessing the effects of a BI as defined by BIEN. Others are testing the effects of programs similar to BI, as defined by BIEN, but with the added feature of a decrease in the BI grant if someone works. I think she refers to this as a guaranteed minimum income.

I suspect that if the U.S. ever did anything like a BI, it would be this guaranteed minimum income version. I think this is because of the vulnerability of a BI, as McFarland defines it, to what I call the “Bill Gates objection”—why give really rich people more money? If one can respond that rich people will not be net recipients because they would pay more in income taxes than they would receive in the BI, this might be a viable response to the objection.

If I am right about this, then studies like the one in Finland, which focuses on a BI, might not tell those of us in the U.S., or in other nations following a similar course, as much as we would hope. That is because the effects of a BI might differ from the effects of a guaranteed minimum income. As an example, if one could get a BI and keep all their earnings without any loss in the amount of their BI grant, such a policy could have a different effect on labor supply than one which would curtail the grant when income from earnings increased. All this means that BI supporters who get enthusiastic about findings from BI experiments ought to take a moment to see if what was studied is what they actually have in mind.

About the author: 

Michael A. Lewis is a social worker and sociologist by training whose areas of interest are public policy and quantitative methods. He’s also a co-founder of USBIG and has written a number of articles, book chapters, and other pieces on the basic income, including the co-edited work The Ethics and Economics of the Basic Income Guarantee. Lewis is on the faculties of the Silberman School of Social Work at Hunter College and the Graduate and University Center of the City University of New York.

Net incomes under a Basic Income system

Net incomes under a Basic Income system

In the May 23, 2017 edition of Basic Income News, Karl Widerquist laments the tendency of some basic income commentators to overstate the cost of a basic income. The typical methodology used to generate these overestimates is as follows:

  1. Obtain the population of the jurisdiction which will implement a basic income
  2. Obtain the amount of the basic income provided to each person in that jurisdiction
  3. Multiply the number in “a” by the number in “b”
  4. The product referred to in “c” is the cost of a basic income

As Widerquist points out, the reason this is an overestimate is that it fails to consider the fact that even though everyone would receive the amount referred to in “b” above, not everyone would be net beneficiaries of this amount.

Suppose the amount referred to in “b” were $10,000, meaning that under a basic income scheme, everyone would receive $10,000 per year. But in every basic income proposal I’ve seen, although the basic income wouldn’t officially be taxed, all other income would be. This means that at some income level, there would be those who’d owe at least $10,000 in their annual tax bill. Since the amount they’d owe in taxes would be at least as large as the $10,000 basic income, they would no longer be net beneficiaries. Their basic income would, in effect, have been taxed back from them. Under a basic income scheme, there would also be those who’d be net beneficiaries of a basic income but not of the full $10,000 amount. All of this might be easier to see if we did a bit of math.

Again, assume that our basic income comes out to $10,000 per year per person. Suppose all other income is taxed at a marginal rate of 25%. The use of one rate is to keep things relatively simple. Here is the key equation for the basic income system being described in this paragraph:

Net Income = $10,000 + (1 – .25) * Other Income

Now let’s play with this equation a bit. Suppose someone had no other income. We’d then end up multiplying $0 by (1 – .25) which would give us $0. And $0 + $10,000 would mean this person would end up with a net income of $10,000. That is, they’d be a net recipient of the full basic income benefit level.

Now consider someone with other income of $30,000. Multiplying $30,000 by (1 – .25), we end up with $22,500. Once we add this to the $10,000 basic income, they’d end up with a net income of $32,500. Let’s look more closely at what’s happened here. The person made $30,000 in other income. If they didn’t have to pay taxes, we’d have (1 – 0), which is just 1, instead of (1 – .25). So they’d keep all $30,000 plus the $10,000 basic income for a net income of $40,000.

Looked at this way, we see that the tax on other income is effectively a tax on the basic income as well. That is, the fact that the person with $30,000 in other income only ends up with a net income $32,500 instead of $40,000 means that $7,500 of their basic income has been taxed back to the government.

Next, let’s take a look at what happens to someone with other income of $40,000.

We’d have to multiply (1 – .25) times $40,000, ending up with $30,000. And $30,000 + $10,000 is a net income of $40,000. If this person paid no taxes on other income, we’d add their $40,000 in other income to the $10,000 basic income for a net income of $50,000. With taxation, their actual net income is $10,000 less than $50,000. That is, we’ve taxed back all $10,000 of their basic income. So this person would no longer be a net recipient of the basic income.

Finally, suppose someone had other income of $100,000. We’d end up multiplying (1 – .25) by $100,000, which comes out to $75,000. Since $75,000 plus $10,000 is $85,000, this person’s net income would be $85,000. Now if they didn’t have to pay taxes, they end up with a net income of $100,000 plus $10,000 or $110,000. But with taxes, their income is only $85,000. We see that not only has their $10,000 basic income been taxed away, so they’re no longer a net recipient of a basic income, but they’re paying enough in taxes to help finance someone else’s basic income, someone with much lower other income than they have.

If we think carefully about these examples, we see what’s wrong with some cost estimates of a basic income: they assume the tax rate in the equation above is 0%. But as I said above, every basic income proposal I’ve seen, assumes that all other income would be taxed at some positive marginal tax rate. This means, of course, that our net income equation will include a term where some positive marginal tax rate will be subtracted from 1. We used a 25% rate for illustration, but really any positive rate will do. This is because any positive marginal tax rate on other income, although not officially a tax on the basic income, is effectively a tax on basic income. And this means some people won’t be net recipients of the benefit. Understanding this point is key to arriving at better estimates of the cost of a basic income guarantee.

 

Addressing uncertainty in basic income

Addressing uncertainty in basic income

Written by: Michael A Lewis

As someone interested in basic income (BI), I read a fair amount about the topic. I read pieces by supporters and opponents, as well as those who might be considered more neutral. I’m often struck by the degree of uncertainty concerning implementation of BI.

A popular argument for BI these days is based on concerns about the possibility of mass technological unemployment. Some in the “tech industry” contend that BI will become necessary as automation replaces more and more human laborers in the years to come. This has led to a debate among economists and others regarding whether automation will result in a net loss of jobs (for humans) big enough to warrant the need for something like BI. Both sides of this debate bring evidence to make their cases. But in the end, we simply don’t know for certain if and when automation will lead to a net loss of jobs for us human beings.

Assuming BI might be implemented in a society which would still require a fair amount of human labor power, we’d like to know what impact BI would have on people’s inclination to sell their labor or, more commonly, “work.” A BI could affect labor supply in at least two ways.

One is that people who received an income they didn’t have to work for may be inclined to work less. The second possible effect has to do with how BI would be financed. If it were financed by an increase in income taxes, this could also reduce labor supply. The reason is that a large proportion of many people’s incomes are earnings, meaning that an income tax is largely a wage tax. A higher wage tax has two possible effects on labor supply.

On the one hand, such an increase could cause people to work less because with the higher tax (and all else equal) their take home pay is smaller than it was before, creating an incentive to work less. On the other hand, a smaller take home pay means one would have to work more than before to maintain their standard of living. This would create an incentive for people to work more not less. If BI were implemented, we have no way of knowing which of these effects would dominate the other.

Leaving the labor market (but still related to it), another area of uncertainty has to do with how people would spend their time, assuming they did reduce their labor supply. Opponents of BI worry that people would use their time “unproductively”, while proponents tend to argue that individuals would engage in more care work or pursue “self-actualization” through pursuing education, writing poetry, starting a business, and the like. But if we’re being honest, regardless of which side of the debate we’re on, we must admit that we don’t have much of an idea what the relative proportion of unproductive to productive activities would be, assuming we could even agree on how to categorize activities as unproductive or productive.

A third area of uncertainty is related to personal relations and household composition. BI could have an effect on who lives with whom, who marries whom, who has kids or not (as well as how many to have), etc. As a society, we obviously differ when it comes to our values about such matters, meaning we might differ on the desirability of BI. But we don’t really know for sure how implementation of BI would affect “family life.”

Now I’m not saying we’re completely in the dark when it comes to questions of BI’s effect on labor supply, use of non-wage time, etc. Economists, sociologists, and others can draw on theory to help us think through these matters. And, by this point, there’ve been several experiments/studies (as well as more recent “startup” studies) which offer a lens on what might happen if BI were implemented. But we should be careful not to overestimate how much help we can receive from such experts, as well as the studies that have been (and are being) conducted.

Considering the many BI experiments (as well as proposed ones) around the world, we need to be cautious about what lessons might be learned. The philosopher Nancy Cartwright, well known for her work in the philosophy of science, has a phrase that’s quite relevant to this discussion: “it works somewhere.” Cartwright frequently utters this phrase within the context of discussing randomized controlled trials (RCTs), the so called gold standard of empirical research in the social sciences. Her point is that even if a well-designed RCT shows that a policy works in one context, that doesn’t necessarily mean it’ll work in another one. This is relevant to BI studies because they’re being conducted, or proposed, in a variety of different contexts. So if we find out that something works in India or Finland, that doesn’t mean it’ll work in Japan or the U.S. In the article cited above, Cartwright goes into great detail about why generalizing experimental findings from one context to another can be so difficult. For those interested in what we might learn from BI experiments, I think her work is quite instructive.

When engineers design systems, such as buildings, bridges, etc., they also must face uncertainties. They don’t know for sure what loads the systems will end up having to bear, they don’t know if there will be earthquakes, they don’t know how forceful the winds will be, etc. One of the things engineers do to deal with such uncertainties is include safety factors in their designs.

For example, suppose an engineer is designing a structure and wind, seismic, and other data indicate that it’ll have to bear a load of 1000 kg. Suppose also that the engineer wants a safety factor of five. Then the load which the structure should be able to bear isn’t 1000 kg but 5×1000 = 5000 kg. So a safety factor is a multiple used to increase the strength or robustness of a system beyond that which is thought to be required to account for uncertainty in what’s thought to be required.

Those of us designing policies don’t have the luxury of being able to use simple equations, which include safety factors, the way engineers do. But perhaps we should adopt a similar safety factor mentality. Implementation of BI would be a complicated undertaking, involving a great deal of uncertainty. Perhaps BI supporters should consider how to increase its robustness in response to labor supply reductions, as well as other unanticipated effects. I admit I’m not exactly sure how to do this. But I believe it’s something worth thinking about.

Michael Lewis

The job guarantee and automation

The job guarantee and automation

Michael A. Lewis

Silberman School of Social Work at Hunter College

In “Why a Universal Basic Income is a Poor Substitute for a Guaranteed Job,” Claire Connelly praises guaranteeing people a right to a job as opposed to guaranteeing them a right to an income. I’ve been involved in quite a few discussions, some of them debates, about the relative merits of basic income versus guaranteed jobs proposals. My position has always been that I have no problem, in principle, with guaranteeing someone a job. If guaranteeing jobs and an unconditional basic income were both financially and politically feasible, I’d be a proponent of both. But if I had to choose one of these policies over the other, I’d prefer the basic income. This is because I think guaranteeing people access to the resources they need to survive has priority over guaranteeing them the right to sell their labor. This, however, isn’t the debate I want to have here. What I want to do, instead, is raise a question about guaranteed jobs proposals: what would it really mean to guarantee someone a job? If those arguing that automation will result in a net loss of jobs for human beings are right, this question becomes especially salient.

Let me start by framing the question more precisely: under what conditions would the government, in its employer of last resort role, hire people? Connelly seems to be supporting the idea of government serving as a buffer stock mechanism. That is, it would step in to hire labor when private sector demand was low and, presumably, step aside when such demand was high. This would put a floor under the price of labor because private sector employers couldn’t, during economic downturns, use the threat of unemployment to get workers to accept lower wages.

As many readers of this site are no doubt aware, some have argued that automation is increasing rapidly and will only continue to do so. Now I’m no expert in this area. So I don’t know if these folks are right that we’re on the path to seeing robots take our jobs. But if they are, this would seem to cause a problem for the buffer stock idea.

The buffer stock approach seems based on a model of the economy where unemployment is due to periodic downturns. The public sector steps in to absorb the resulting labor surplus, but this is meant to be temporary. Once the economy starts growing again, and unemployment declines, public sector employment can contract, as those who worked in the public sector are absorbed by the private one. But automation isn’t supposed to work like that. Instead, there is a steady, but permanent, decline in many types of work as machines take our jobs. I know there’s a huge debate about whether other types of jobs will develop to replace the one’s lost to robots: but suppose such replacement doesn’t happen. Perhaps it’s premature to do so, but I wonder if guaranteed jobs proponents have thought about how their guaranteed jobs plan would work in such an environment. Would government indefinitely hire all those who’ve lost their jobs to machines?

On the ‘automation’ argument for basic income

On the ‘automation’ argument for basic income

Written by: Michael A. Lewis
Silberman School of Social Work at Hunter College and the CUNY Graduate Center

When I first became interested in the basic income, I was a graduate student studying welfare reform. For those who aren’t in the know, “welfare” is the more common name used in the U.S. to refer to a program called Temporary Assistance for Needy Families (TANF) and which used to be called Aid to Families with Dependent Children (AFDC). TANF and AFDC aren’t exactly the same programs, but they do have some key things in common: they provide financial support to low income persons, most of those who receive such support are women and children, and, I think it’s fair to say, both programs are somewhat controversial.

The controversy around welfare has to do with the fact that many of those who receive benefits are apparently “able-bodied” persons who’re thought capable of working (“working” in this context means selling one’s labor in return for a wage, instead of, say, taking care of one’s children, something many would regard as work). Yet not enough of those on welfare are working, according to a common belief among many U.S. citizens/residents as well as, apparently, politicians. So in an attempt to socialize welfare beneficiaries into understanding the importance of work, many of them are required to work in return for their benefits, a practice commonly called “workfare.” Many also remain poor, even after receiving benefits, because the financial support they receive is pretty meager.

As a graduate student, I thought workfare, as well as the low level of benefits provided to recipients was a very unjust way of assisting poor persons; I also thought we could do better (in fact, I still think these things). My entry into the world of basic income was because I believed it a more just way of addressing poverty than welfare and related programs.

Once I started studying basic income and meeting others interested in the idea, I heard other justifications for it. It would enhance freedom, it would allow people to engage in care work if they so choose, it would give people an income representing their share of commonly owned natural resources, it would be a way of replacing some or all of the welfare state (which, of course, assumes there is something wrong with the current system), etc. But the argument that seems to have caught on the most, at least in the U.S., is the idea that a basic income will become necessary as robots/machines take our jobs.

I have to admit that part of me has been a bit concerned about the degree to which the automation argument seems to dominate basic income discussions. My worry is that as we spend so much time debating who’s right about whether robots will take most, or perhaps all, of our jobs and, therefore, whether there’ll be a need for a basic income, other arguments for such a policy get “crowded out” of the discussion. Yet as I’ve voiced this concern, mainly to myself, I’ve also wondered why this argument for a basic income seems to have caught on in a way that others haven’t?

I think part of the answer has to do with where I started—U.S. citizens/residents worry a lot about the degree to which healthy people work to take care of themselves (and their families) and are quite skeptical about policies they believe will allow people to shirk this responsibility. But I think another part of the answer has to do with the role of race in our society. I suspect that in the minds of many citizens/residents the degree to which a basic income would allow people to shirk their obligations to work would vary by race. To put it bluntly, I suspect many assume that black and brown people would be more likely to shirk this responsibility than whites would be. If I’m right about all this, then perhaps it shouldn’t be surprising that the U.S. isn’t naturally the most fertile place for the basic income idea to take hold. But why would it take hold in the form of the automation argument? I think the answer here might be pretty simple. If machines are about to take all our jobs, then automation represents a relatively indiscriminant force. That is, “hard working white people” might be threatened just as much as “lazy shiftless brown ones” are. Perhaps this has been enough to get white folks to take notice of a policy that perhaps could address the problem.

About the author: Michael A. Lewis is a social worker and sociologist by training whose areas of interest are public policy and quantitative methods. He’s also a co-founder of USBIG and has written a number of articles, book chapters, and other pieces on the basic income, including the co-edited work The Ethics and Economics of the Basic Income Guarantee. Lewis is on the faculties of the Silberman School of Social Work at Hunter College and the Graduate and University Center of the City University of New York.