Modeling the Macroeconomic Effects of a Universal Basic Income – Rakeen Mabud and Felicia Wong. * A universal basic income would grow the economy – Dylan Matthews.

Our colleagues at the Roosevelt Institute, together with the Levy Institute, just published an exciting new paper entitled, “Modeling the Macroeconomic Effects of a Universal Basic Income.”

The paper takes a major step forward in answering an important question: How would a massive federal spending program like a “universal basic income” (UBI) affect economic growth and economic inequality? We live in a time when many big new economic policy ideas are suddenly on the table—from unconditional cash assistance programs like UBI; to single-payer health care, which would have the federal government as our sole medical insurance system; to significant increases in both individual and corporate taxes. At Roosevelt, we do our best to use data and analysis to understand who would benefit from new policies, who would lose, and how the economy as a whole—our institutions and the power relationships among them—would change as a result. So it is exciting to see our report on UBI, one of the biggest proposed game-changers, generate thoughtful media coverage. 

All models, of course, have baked-in assumptions, which are important to unpack. So, the rest of this post explains a little more about what those assumptions are; why many Roosevelt economists and researchers believe those assumptions have merit; and why we are proud members of the UBI research community. But we should note up front: We view this as part of a much larger conversation about UBI and other associated questions, including our assumptions about just how much the economy still has room to grow. Our paper shouldn’t be read as pro- or anti-UBI. We believe strongly in the goals that many UBI adherents have: the importance of alleviating poverty and suffering, the importance of giving workers and other low-income people more power and greater tools for self‑determination, and the importance of economic security for all Americans. But our position is that we still have a lot to learn about UBI, cash assistance, and improving the social safety net.

The Levy model and what we tested

The paper explores the macroeconomic effects of a UBI using the Levy Institute’s macroeconomic model. Important note: Economic models are simply stylized approximations of how our economy works, and we should be careful not to conflate what the model says with reality. The Levy model was developed over the course of years by comparing the model’s predictions to what actually happened in the real world. Through this iterative process, each round honing in on a “truer” picture of the economy, the team at Levy developed a baseline forecast—a prediction of what would happen to important components of the macro economy (e.g. household wealth and income and the government’s net debt) if nothing else changed and the Congressional Budget Office’s overall GDP forecasts came true. By comparing the effects of different sized UBIs to the Levy baseline, we can roughly predict how huge amounts of federal spending would affect inflation and other macroeconomic outcomes.

In the paper, we examine three versions of a UBI: $1,000/month for all adults, $500/month for all adults, and a $250/month child allowance. The headline result is that the model predicts that if we funded a UBI without raising taxes, GDP would grow—an eye-catching 12.56%, in the case of the full $1,000/month UBI. The model also predicts that even in tax-financed UBI scenarios, the economy would grow (because we incorporated the assumption that an extra dollar going to a poorer household is more likely to be spent rather than saved, something wonks call “the marginal propensity to consume.”) Dylan Matthews at Vox did a great job of explaining all of the results, so we won’t go further here, but it is worth reading his piece and the paper itself for the detail.

Economic worldview, and why it matters

We did want to emphasize the big-picture context here, because it’s important. The Levy model, and the results it offers us about the macroeconomic effect of a UBI, is situated in a larger worldview that many of our Fellows, including Mike Konczal and Marshall Steinbaum, hold: that there’s a lot of room to grow in our economy. Crucially, this is a Keynesian view, assuming that total spending is what determines economic output. As such, the model holds that policies that would increase total spending, like UBI, will grow the economy. The model also makes two other important assumptions that adhere to the broader perspective that the economy is not operating near potential output. First, unconditional cash transfers do not mean that people stop working, which our recent paper by Ioana Marinescu, surveying extensive empirical evidence, demonstrates. Second, the paper takes seriously the idea that when the government takes away an additional dollar of a rich household’s income through taxes, the household doesn’t respond by dropping out of the labor force. Instead, we believe that each marginal dollar taxed away reduces rich households’ incentives to bargain for a larger share of the economic pie. Our belief about this “bargaining elasticity” derives from important work by Thomas Piketty, Emmanuel Saez, and Stefanie Stantcheva, who show that the bargaining effect trumps the labor supply effect—it isn’t even close. This means that high progressive tax rates could have a very positive effect on reducing inequality, without harming growth. This is a big assumption, and we fully recognize that.

Implications and today’s economic debate

The fundamental idea that there’s a lot of room to grow in our economy, and as a result, a lot of room to think bigger about our social safety net, has implications for problems far beyond what a UBI is designed to address. What is the optimal interest rate? What is the best trade off between inflation and unemployment? What are the most effective ways to increase long-stagnant wages? The answers to each of these questions stem from one’s view of whether the economy is operating at its fullest capacity.

Further, the UBI conversation happening right now is important because it forces us to tackle big questions head on: If we can do more to expand the social safety net, what should that expansion look like? Who should get that expansion? Workers? What does it mean to work? Does going to school or raising a child count as “working”? Who is deserving of government support? We are privileged to be a part of the UBI community—one that enthusiastically tries to answer these very big questions, and, as a result, ends up breaking past all kinds of constraints in the way we think about how our economy and society are structured.

Of course, today’s political destabilization produces deep anxiety, and indecisions, or bad decisions, on the part of our current political leadership has caused real suffering in the lives of real people. But in one respect, our political era is liberating. We are now talking not just about a guaranteed cash income, but also about job guarantees (the idea that everyone who wants to work will have paid work to do); baby bonds (wherein all children, at birth, would be given a bond that would come due when that child became an adult—and thus all, regardless of their family wealth, would have some economic security); much higher tax rates on the wealthy; and innovative ways to break up monopoly companies that may be choking our economy and our workers: These are all real conversations today. And they are all radical ideas that embrace the concept that we as a society are able to solve big problems. We hope that this paper contributes to this ongoing conversation, and serves as one more piece of evidence that thinking bigger is not foolhardy, but rather, imperative.

Download the report: Roosevelt


Study: a universal basic income would grow the economy – Dylan Matthews.

A universal basic income could make the US economy trillions of dollars larger, permanently, according to a new study by the left-leaning Roosevelt Institute. 

Basic income, a proposal in which every American would be given a basic stipend from the government no strings attached, is often brought up as a potential solution to widespread automation reducing demand for labor in the future. But in the meantime, its critics typically allege that it is far too expensive to be practical, or else that it would spur millions of Americans to drop out of the labor force, wrecking the economy and depriving the government of a tax base for funding the plan.

The Roosevelt study, written by Roosevelt research director Marshall Steinbaum, Michalis Nikiforos at Bard College’s Levy Institute, and Gennaro Zezza at the University of Cassino and Southern Lazio in Italy, comes to a dramatically different conclusion. And it does so using some notably rosy assumptions about the effects of large-scale increases to government spending, taxes, and deficits, assumptions that other analysts would dispute vociferously.

Their paper analyzes three different models for a universal basic income:

  1. A full universal basic income, in which every adult gets $1,000 a month ($12,000 a year)
  2. A partial basic income, in which every adult gets $500 a month ($6,000 a year)
  3. A child allowance, in which every child gets $250 a month ($3,000 a year)

They find that enacting any of these policies by growing the federal debt — that is, without raising taxes to pay for it — would substantially grow the economy. The effect fades away within eight years, but GDP is left permanently higher. The big, $12,000 per year per adult policy, they find, would permanently grow the economy by 12.56 to 13.10 percent — or about $2.5 trillion come 2025. 

It would also, they find, increase the percentage of Americans with jobs by about 2 percent, and expand the labor force to the tune of 4.5 to 4.7 million people.

They also model the impact of the plan if it’s paid for with taxes. That amounts to large-scale income redistribution, which, the authors argue, would stimulate the economy, because lower-income people are likelier to spend their money in the near-term than rich people are. Thus, they find that a full $12,000 a year per adult basic income, paid for with progressive income taxes, would grow the economy by about 2.62 percent ($515 billion) and expand the labor force by about 1.1 million people.

These are extremely contentious estimates, borne of controversial assumptions about the way the economy works and the effects that a basic income would have on it. Many, if not most, economic modelers would come to very different conclusions: that a basic income discourages work, that raising taxes to pay for it could have profound negative economic impacts, and that not paying for it and exploding the deficit is a recipe for fiscal and economic ruin.

But the authors argue that the economic model they’re using, run by the Bard College Levy Economics Institute, uses more realistic assumptions than alternative models, and is particularly well-suited for predicting a UBI’s impact.

If nothing else, the paper should provoke other analysts with less rosy models to come up with their own predictions of what a large-scale basic income would do.

The predictions driving the model

Predicting how policies will change the economy is difficult, and requires making tricky and contestable assumptions about the way the economy currently works. And the Levy Institute model and Roosevelt Institute researchers make some big assumptions that many macroeconomists and public finance economists are likely to disagree with.

Perhaps the single most important assumption is that the economy is suffering from a lack of demand — consumers and businesses aren’t buying enough stuff. This is traditionally the problem in recessions, and it’s typically addressed through monetary or fiscal policies meant to boost demand. Governments can spend a bunch of money on infrastructure or tax breaks, which juices demand by having the government buy more or giving consumers more money to spend; this was the approach of the 2009 stimulus package. Or central banks can print a bunch of money and bring down interest rates, which makes it easier for businesses and individuals to borrow and spend money.

But in recent years there’s been growing concern that this might be a “secular” problem — one that persists even after the economy has seemingly recovered from a downturn. Larry Summers, the former Obama adviser, Harvard president, Treasury secretary, and eminent economist, has since 2013 warned of an ongoing “secular stagnation”, borrowing the term from the early 20th century economist Alvin Hansen, who coined it in the 1930s. Thomas Piketty’s book Capital in the 21st Century has prompted research tying this decline in aggregate demand to the surge in high-end inequality, if the rich control more and more money, and aren’t spending it, that helps explain sluggish demand.

The Roosevelt Institute has done a bunch of work on this theme, arguing that a shortfall in aggregate demand, tied to inequality, is strangling the economy. Roosevelt’s JW Mason argued in a June 2017 paper that the US economy is still quite far from full employment, and GDP is falling below its potential level, largely because of low demand; he argues for more aggressive Federal Reserve policy to correct the problem. Roosevelt’s Steinbaum and Mike Konczal argued in 2016 that low levels of demand have led to less labor mobility (fewer people switching jobs and moving for work), less entrepreneurship, and more concentration of profits in a handful of companies at the expense of competitors and potential new challengers.

So the Levy model used in the basic income paper, building off this research, assumes that demand is well below where it could be. That helps explain why big, unfunded increases in spending, like a basic income not funded by taxes, could be stimulative. Many economists would agree that in, say, 2009, when unemployment is surging and the economy’s in recession, it makes sense to do big deficit spending to get demand up again. The Levy model is arguing that a demand shortfall is still a problem, and similar measures would thus be effective in 2017 just as they were in 2009.

Plenty of other models would come to different conclusions. The Penn Wharton Budget Model, a widely used economic forecasting model used by politically unaligned groups like the Tax Policy Center, tends to project that policies that increase the deficit dramatically will have significant economic downsides. For instance, when the Tax Policy Center looked at Donald Trump’s likely tax plan, it used the Penn Wharton model and found that over two decades the tax plan would reduce GDP by 4 percent, mostly because the increased deficit would lead to a surge in interest rates on government debt, and then on small business loans, mortgages, and credit cards as well.

The Roosevelt-Levy model, by contrast, finds that any negative effect due to the deficit would be swamped by the positive economic impacts of increased demand.

The Roosevelt-Levy model also makes two other significant assumptions:

  1. A basic income does not discourage work at all
  2. Households don’t respond to changes in their tax burden

These are both contentious, but reflective of Roosevelt’s broader view of how the economy works. University of Pennsylvania economist Ioana Marinescu recently conducted a wide ranging review of the literature on unconditional cash programs for Roosevelt, focusing on programs in the US and Canada. She examined experiments in the 1970s and ’80s that evaluated “negative income taxes” (NITs, essentially basic incomes that phase out as you earn more), Alaska’s Permanent Fund (which taxes oil extraction and returns the money directly to every man, woman, and child through an annual check), and a dividend the Eastern Band of Cherokees issued to members of the tribe from casino revenues.

All of these cases find reductions in work that are, at most, modest. In the Cherokee case (where members got about $4,000 to $6,000 a year) there was no effect on work; in the Alaska case, where checks are generally $800 to $2,000 per person (so up to $8,000 for a family of four), there’s a small increase in the share of people working part time, but no overall effect on the share of the population working. Indeed, the part-time work boost could come from people entering the workforce anew. The negative income tax experiments are more complicated. Most of the studies found no statistically significant reduction in work; only one, the Seattle/Denver experiment, found a reduction, and it saw the employment rate fall by 4 percentage points

In any case, Marinescu concluded that the effects on work are small but minimal. “Our fear that people will quit their jobs en masse if provided with cash for free is false and misguided,” Marinescu writes. 

The new Roosevelt paper cites Marinescu’s finding as support for its assumption that there’d be no decline in work. But even if there’s a modest decline in work — or a socially salutary decline, due to new mothers taking more leave or kids staying in school longer or people choosing to retire earlier because they want to — that could have significant economic effects, which the model will not pick up.

The assumption of no household response to changes in taxes is also sure to be contentious. It cuts two ways. On the one hand, most conservative/libertarian leaning economists would argue that taxes, especially corporate and investment taxes and high individual income tax rates, have a tremendous impact on business decisions and individual decisions about work. This is the whole idea behind supply-side economics as practiced by the Reagan, Bush, and most recently Brownback administration in Kansas: that lowering taxes on economic activity will lead to more of it, which grows the economy. Correspondingly, enacting big tax increases to pay for a basic income would be predicted to hurt the economy.

The right-leaning Tax Foundation, for instance, has an economic model that predicts huge positive impacts from cutting taxes on corporations and high-income individuals, and significant negative impacts from raising them. For instance, the Tax Foundation found that the House Republican plan to cut individual and business taxes would lead to a 9.1 percent boost in GDP And a 7.7 percent boost in wages, and that Bernie Sanders’s tax plan WOULD CUT GDP by 9.5 percent and would lower the average American’s income by nearly 13 percent.

On the other hand, some left-leaning economists have argued that raising taxes on high earners has positive effects on household behavior. Berkeley’s Emmanuel Saez, Piketty, and Harvard economist Stefanie Stantcheva have argued that very high marginal tax rates (the top rate on wages was 91 percent for most of the 1950s) discourage the rich from making very large salaries. In particular, it prevented them from bargaining with their employers to divert money from shareholders or lower-ranked staffers into higher executive compensation.

Think of it this way. In 2017, the top federal income tax rate is 39.6 percent. So if a CEO convinces his company to raise his pay from $5 million to $6 million, he’ll get to keep $604,000 of that raise (I’m ignoring state and payroll taxes for the sake of simplicity). That’s a really healthy after-tax raise, so that CEO has a very big incentive to lobby for pay hikes like that. But in 1955, the top federal income tax rate was 91 percent. So that same pay raise would only net him $90,000. Not nothing, but a way, way smaller windfall. So back then, executives had less reason to try to fight to earn more, which kept down inequality.

The Roosevelt paper predicts that the super-negative Tax Foundation story, where big tax increases translate into economic calamity, would not come to pass, nor would the rosy Saez/Piketty/Stantcheva story, in which big tax increases translate into a quite big reduction in inequality, which in turn could alleviate secular stagnation by putting more money in the hands of the middle-class to spend.

This is predicting a huge, huge policy change

It’s worth being clear about just how significant some of the policies envisioned in this paper are.

Some of them are relatively modest in cost. A $3,000 per child per year child allowance is well in line with what other rich countries do; Canada gives family with children under 6 about US$5,000 a year, and families with older children about US$4,250 (the benefit phases out for the rich). That plan would be totally affordable for the US with relatively small tax increases, especially if we replaced other income support programs for children. A group of influential US property experts have recently recommended a $3,000 to $3,600 a year child benefit, and the Child Tax Credit Improvement Act from Congress member Rosa DeLauro, legislation backed by Nancy Pelosi and John Lewis, would create a new $3,600 fully refundable tax credit for kids under 6. This is well within the American political mainstream, and it could cut child poverty by a quarter, a third, or even in half.

But a $12,000 a year per adult basic income is another matter. The Roosevelt paper finds that paying for it would require increasing household tax revenue by 120 percent — a more than doubling. To do that, it assumes that all but the poorest 40 percent pay more in taxes. The middle quintile (households with an income between $48,300 and $85,600, per the Tax Policy Centre) would see their average tax rate go from 14 percent to 25 percent; the top one percent would see its average tax rate go from 32.9 percent to 67.9 percent.

This would be historically unprecedented. Note that even in 1979, when the top marginal tax rate was 70 percent, the top one percent paid only 35.2 percent of their income in federal taxes on average. To nearly double that number, we’d need much, much higher income tax rates, and likely a few other new measures to tax the rich.

Of course, implementing a basic income and not paying for it would not have these problems. But it would explode the deficit so dramatically that even many left of center economists would consider it dangerous.

All of which is to say, the most ambitious proposals that the Roosevelt paper considers would need to be refined considerably to be ready for implementation. There are big tradeoffs between growing the deficit and forcing big tax increases on the rich and middle class alike, and it may be that the child allowance, which Roosevelt predicts would do the least for the economy, is the most fiscally viable option.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s