What Is Basic Income?
Despite significant federal spending on anti-poverty programs, poverty and economic insecurity remain defining features of American life.
The rapidly shifting economy of the past 50 years has decoupled economic progress from shared prosperity. Power asymmetries between employees and employers, stagnating wages, wealth inequality, and the specter of job automation all threaten to leave more Americans behind in the coming decades.
A basic income is an unconditional, regular cash payment that guarantees a baseline of economic stability to all citizens.
Despite significant variation in specific proposals, all basic income programs share four key characteristics :
Periodic: a recurring payment, as opposed to a one-time grant.
Cash: the payment is strictly cash, as opposed to in-kind services.
Individual: the payment is made to individuals, not households.
Unconditional: the only eligibility criteria to receive the payment is citizenship.
Amidst continuing poverty, mounting inequality, and impending volatility, basic income offers a way to both strengthen and raise the safety net, providing a stronger foundation of income for all American citizens, no matter what.
Despite the appeal of basic income and its simplicity, and mounting positive evidence from small experiments, big questions remain. Is basic income the best use of government funds, or would alternative investments prove more effective? How much is “basic”, and how best to offset the program’s cost? How might a national scale program differ from the experimental evidence from small-scale experiments?
Proposals differ on:
Whether the payments are "universal" to all citizens, or "targeted" towards those with lower incomes.
The amount of income guaranteed by the program.
Whether guardians receive additional payments for children.
How much of the existing welfare safety net it replaces.
How to offset costs.
Depending on its design, advocates argue a basic income could:
Significantly reduce or eliminate poverty.
Stimulate growth by redistributing cash towards those who are more likely to spend it.
Promote entrepreneurship by decreasing risk.
Increase the bargaining power of workers and reduce inequality.
Support a series of social benefits, including improved physical and mental health, educational and familial outcomes, and reduce crime.
Decrease the bureaucratic inefficiencies of welfare provision by combining patchwork income support programs into a single program.
Skeptics argue that a basic income could:
Discourage work and foster welfare dependence.
Be prohibitively expensive and irresponsibly spent.
Be less effective than other uses of tax revenue.
Cause price inflation.
What does the evidence say?
Basic Income on Growth
Current evidence suggests a number of pathways basic income might grow the economy:
- Putting more money in the hands of those most likely to spend it
- Reducing the risks of entrepreneurship
- Enabling educational and professional investments
- Reducing other welfare expenditures by lifting citizens from poverty
But to determine the overall impact, the benefits of unconditional cash must be weighed against the potential costs of taxes used to fund the program.
While current evidence cannot answer all questions about basic income, current findings strongly counter two common concerns. Receiving basic income does not discourage work, and provisioning basic income does not cause inflation.
Basic income could grow the economy by increasing demand and raising long-term earnings through educational and professional investments.
Whether basic income increases aggregate demand may ultimately depend upon the balance between the redistribution of spending power on one hand, and the effects of taxes used to fund it on the other.
By redistributing money from the the wealthy to lower-income citizens, basic income puts more money in the hands of those who are more likely to spend it . While this would raise aggregate demand across the economy, the potentially countervailing effects of higher taxes must be considered to determine overall impacts .
In addition to increased demand, basic income may also lead to educational and professional investments that lead to higher wages in the medium to longer run . For example, in California’s SEED pilot program, receiving basic income enabled recipients to work fewer part-time shifts in favor of finding better employment, complete training programs and internships, and even take greater risk in search of better opportunities .
In the longest-run study of cash transfers to date, Mexican infants whose mothers received conditional cash transfers show significant benefits in education, wages, and geographic mobility twenty years later. And during a series of basic income experiments in the U.S. in the 1970’s, receiving basic income was associated with an increase in the number of adults pursuing higher education, and higher high school completion rates .
Research on the impact of expanding the child tax credit to a child allowance - effectively implementing a basic income for kids - found that the program’s cost of $100 billion per year would be offset by $865 billion in annual monetary and non-monetary benefits . These include $80 billion per year from children’s increased future earnings, and $670 billion from improvements in child health and longevity.
In developing countries, where cash transfer programs are externally funded (without taxation, often through charity), the economic effects are significant and positive on growth. In Namibia, basic income raised the rate of those engaged in “income generating activities” from 44% to 55% . In Mexico, basic income recipients invested 26 cents of every peso in productive assets, raising agricultural income by nearly 10% and raising long-term consumption by 1.6 cents . In Kenya, a one-time cash transfer of $1,000 to over 10,500 poor households led to sharp increases in income and consumption , with roughly 2.6 dollars worth of increased economic activity per dollar spent (despite the sudden spike in money supply, minimal price inflation occurred).
Econometric simulations of the impacts of a nation-wide program produce conflicting results depending on the model’s assumptions. The Roosevelt Institute, using the Levy Model, found that a tax-financed universal basic income of $6,000 per year would increase GDP by 1.65% , and a $12,000 per year UBI would increase GDP by 2.62% . However, the Penn Wharton Budget Model, built with different assumptions, predicts that a tax-financed UBI of $6,000 per year would decrease GDP by 1.7% .
A robust body of research finds that receiving unconditional cash does not create dependency or discourage work. However, existing experiments are limited in what they can tell us about the effects of a full-scale, national program. Few experiments have studied higher-level payments in the US.
A 2020 systematic review of 38 empirical studies on the relationship between unconditional cash transfers and basic income found no evidence that receiving unconditional cash transfers creates dependency or discourages work :
Regarding concerns that unconditional cash transfers discourage work, The Stanford Basic Income Lab writes: "The evidence from existing UBI related schemes, however, negates many of these concerns and, overall, indicates that UBI-related programs have marginal effects on labor market participation."
The results are especially clear in the developing world. As Nobel laureate Abhijit Banerjee and colleagues conclude in their aptly titled paper, Debunking the Stereotype of the Lazy Welfare Recipient: "We...find no systematic evidence that cash transfer programs discourage work."
Contra such concerns, monthly recipients of an unconditional $500 in the Stockton Economic Empowerment Demonstration (SEED) obtained full-time employment at more than twice the rate of non-recipients .
Research on the child tax credit - a basic income for the families of children - finds similar outcomes. When president Biden expanded the existing Child Tax Credit (CTC) to an unconditional program with no work requirements, there was no change in the employment rates of recipients . Similarly, research on expanding the CTC finds that a $1,000 increase in the benefit amount is associated with a 0.37 hour , or 1.1% in working rates of recipient single mothers. In line with the SEED basic income experiment, receiving unconditional cash may in fact increase, rather than decrease, employment rates.
One notable exception is a series of negative income tax (NIT) experiments conducted in the 1970’s across the US and Canada. While initial results found significant reductions in work-participation for NIT recipients, subsequent studies discovered errors in their methodology. Accounting for errors, as well as self-reporting bias, the findings suggest minimal work-participation reductions, in line with findings from more recent studies.
Econometric models - across all assumptions - predict similarly negligible impacts of a nation-wide basic income on employment. The Levy Model predicts a 0.31% increase in the employment rate , while the Penn Wharton Budget Model predicts a 3.2% reduction in hours worked by households (roughly 1.2 hours).
Even if a nation-wide basic income did lead to modest reductions in overall working hours, standard economic theory suggests those hours could be filled by presently unemployed job-seekers, tightening the labor market and improving the overall efficiency of the economy .
Basic income could increase entrepreneurship and small business formation by decreasing the risks of failure.
The share of US employment accounted for by young firms has declined nearly 30% over the past 30 years. A growing body of evidence suggests that more generous safety nets lead to higher rates of entrepreneurship. One rationale is that strong safety nets reduce the risks of failure faced by entrepreneurs.
For example, studies on expansions of both public health insurance and food stamp eligibility found increases in self-employment by 15% , and new firm incorporation by 12% , respectively. Notably, in the case of expanded food stamp eligibility, Walter Frick comments that many of the newly eligible entrepreneurs didn’t enroll in the expanded food stamps program. Instead, "simply knowing that they could fall back on food stamps if their venture failed was enough to make them more likely to take risks."
In France, an expansion of unemployment benefits to allow those transferring from unemployment to self-employment to temporarily continue drawing benefits, and guaranteeing that should their venture fail, they would once again be eligible for unemployment benefits, led to a 25% increase in new firm creation .
Research on the Alaska Permanent Fund (APF) finds that receiving unconditional dividends boosted entrepreneurship rates , but the effects dissipated over time. The study’s authors speculate that one reason may be the gradual decline in the dividend size.
While previous studies have established a negative relationship between overall government spending and entrepreneurial activity, one study suggests that what matters is not the overall size of government, so much as the composition of its spending. Given a fixed pie of government spending, increasing budget allocation towards social and public goods (of which basic income could be an example) relative to private subsidies (such as to agriculture, energy and fuel, manufacturing, defense spending, and mining), is associated with higher entrepreneurial activity .
However, there has been little direct study on how basic income programs may impact entrepreneurship, in part due to the limitations inherent in small pilot programs. Thus, the question of whether basic income is the most effective expansion of government programs in terms of entrepreneurship relative to other possibilities, like universal healthcare, job guarantees, or expanding existing programs, remains unclear.
By consolidating various welfare programs into a single basic income, we could reduce bureaucratic inefficiencies while more effectively reducing poverty.
Despite $361 billion of federal spending on safety net programs in 2019, 34 million Americans remained in poverty .
The U.S. has a variety of overlapping welfare programs, each with different eligibility criteria, benefit amounts, and administrative infrastructures. As a result, inadvertent “poverty traps” are created, disincentivizing low-income workers from increasing their incomes , lest they lose more in benefits than they gain in wages.
Instead, the multiple income support welfare programs could be consolidated into a single basic income, reducing bureaucratic inefficiencies and poverty traps, while more effectively reducing poverty .
In European econometric simulations, basic income (both negative income taxes and universal basic income) performs better than existing systems of conditional welfare in terms of maximizing a social welfare function .
Similarly, in such models, if redistribution does not discourage work (against orthodox assumptions, but in line with current research), then unconditional, negative income tax style programs better optimize social welfare functions than programs with work requirements like the existing earned income tax credit (EITC).
By lifting citizens from poverty, basic income would reduce public spending on healthcare, incarceration, and housing.
Money spent on lifting citizens from poverty often has a fiscal multiplier effect, where every dollar spent returns more than a dollar’s worth of savings.
For example, a Vancouver experiment gave a one-time transfer of $7,500 to recently homeless individuals. Compared to a control group, the cash recipients saved an extra $8,172 per year of public expenditure via spending fewer nights in homeless shelters.
Research on the potential effects of a child allowance - a basic income given to children’s families - found that $100 billion of annual spending on the program would return $865 billion in annual benefits, the majority of which comes from improved health and longevity. In particular, the public receives $3.5 billion per year in savings from reduced expenditures on children’s and parents’ healthcare costs .
In North Carolina, parents receiving an unconditional $4,000 per year reduced arrests for minors by 22%.
There is no evidence - nor theory - that suggests a tax-financed basic income would cause price inflation.
As the Stanford Basic Income Lab writes:
The case of Alaska provides a large natural experiment. The Alaska Permanent Fund Dividend (APF) is an annual, unconditional payment given to all Alaskan citizens since 1981, generally between $1,000 and $2,000 per year. Following the introduction of the APF, Alaska's inflation levels actually declined. In the four decades since, Alaska has experienced lower inflation levels than the United States average .
Additional experiments, from Kuwait granting every citizen $3,600 in celebration of the country's 50th anniversary, to experiments across Mexico, Kenya, and India similarly find no inflationary effects associated with cash transfers.
Neither do econometric models that predict the impacts of larger UBI programs find significant inflation. The Levy Model predicts that even a deficit financed - the most likely funding method to spike inflation - UBI of $12,000 per adult would only raise price levels by 3.77% over 8 years. Notably, a simulation of UBI in New York City found that housing prices would actually decrease .
Basic Income on Stability
In terms of education, physical and mental health, crime, family relationships, poverty, and inequality, the observed effects of basic income are overwhelmingly positive.
This should come as little surprise. All else equal, giving citizens more money should produce positive outcomes. Still, the question remains as to whether basic income is the best approach relative to other possibilities, and how high a payment is necessary to achieve optimal benefits.
The effects on poverty and inequality depend significantly on the benefit amount, the financing method, and how much existing welfare basic income replaces. While a basic income tied to the poverty line could eliminate poverty, replacing all other welfare programs would leave the most vulnerable populations worse off.
A basic income set at the poverty line would eliminate federally recognized poverty in the U.S. Smaller programs would still significantly reduce poverty.
In 2021, 82% of Americans were dissatisfied with the nation’s efforts to reduce poverty and homelessness .
A basic income that eliminates poverty could be achieved in at least two ways: a universal basic income with an annual payment equal to the poverty line, or a negative income tax with an income floor set at the poverty line, which then phases out as earned income rises.
Using data from 2004, one estimate suggests that a negative income tax that fully eliminated poverty would’ve cost between $219 - $336 billion (in 2007 dollars), depending on the phaseout rate.
Smaller basic incomes could still substantially alleviate poverty. A small basic income of $250 per month ($3,000 per year) given in full to all households who earn less than $150,000 per year is estimated to reduce poverty by 45%, and childhood poverty by 65% .
Similar estimates find that raising money with a 10% value added tax and disbursing the revenue to all citizens in the form of a UBI set at 20% of the poverty line, or $2,500 per year, would progressively benefit the bottom 60% of the population in terms of income .
Looking to evidence from Alaska’s Permanent Fund Dividends, findings on inequality are mixed. Some research finds that distributing the unconditional payment to all citizens reduced inequality . Others found that in the long run, it may increase inequality if wealthier recipients invest the dividend in higher yield assets , while lower-income recipients spend the payment on goods and services.
This suggests that basic income’s effects on inequality may depend on the tax structures used to offset the cost.
Replacing all other welfare programs with basic income would leave the poor, parents, the elderly, and the disabled worse off.
Varieties of basic income proposed by Milton Friedman, or more recently Charles Murray, are conceived as full replacements for the entire welfare state. This can be described as a “pure UBI”. A revenue-neutral swap of the welfare state for an equivalent basic income would disproportionately benefit the childless, young, and non-disabled , thus benefitting the middle-class more than those most vulnerable or with the lowest incomes.
The same pattern prevails in the UK. A UBI that replaces all other means-tested benefits would benefit the middle-class more than lower-income groups , and would increase poverty rates for children, pensioners, and working age adults .
But when most other programs are held constant, UBI may be preferable to certain existing programs. For example, studies comparing UBI against both unemployment and the earned income tax credit find potential advantages .
In terms of where to draw the line regarding what should remain, and what should be replaced, two heuristics prove helpful. First, basic income is a more effective program in well-functioning markets. In areas with underlying market problems, whether information asymmetry, externalities, or barriers to entry, basic income is a poor tool .
Second, the distinction between welfare programs that serve “insurance” vs. “income support” functions. In theory, basic income could replace all income support welfare programs, to the benefit of recipients and non-recipients alike .
Basic income recipients report significant gains in mental health.
Basic income may significantly improve mental health by raising the baseline of economic security. Beyond material deprivation, poverty is a matter of psychological deprivation, driving psychological feedback loops that perpetuate poverty .
In both the employed and unemployed, expectations of future economic insecurity significantly reduces mental health (up to 0.18 standard deviations from the mean) . These future expectations of economic insecurity are more detrimental to mental health than actual, realized episodes . In particular, fear of unemployment is the most damaging form of economic insecurity . By raising the income floor, a basic income could reduce the depths of perceived future economic insecurity.
In California, results from the SEED experiment find that basic income recipients experienced clinically significant gains in mental health . In Kenya, a one-time unconditional transfer equivalent to $1,076 had greater positive impacts on psychological wellbeing than a five week psychotherapy program . In India, low-income workers receiving extra cash transfers saw mistake rates decline and productivity rise by 6.2%, which the study authors attribute to the psychological benefits of reducing financial concerns .
Some studies even find that by reducing mental stresses, unconditional cash transfers reduce addictive behaviors such as alcohol and drug use, and gambling .
In Finland, basic income recipients over a two year period reported improved mental health and cognitive function. In particular, recipients reported:
And finally, a meta-analysis of cash transfers on self-reported mental health found significant benefits in low and middle income countries , with an average effect size of 0.1 standard deviations.
Receiving basic income is associated with improvements in health, reductions in crime rates for minors, higher educational attainment, and improved familial relationships.
In Dauphin, Canada, recipients of basic income had an 8.5% decrease in hospitalization rates relative to control groups. In Alaska, each additional $1,000 in unconditional dividends from the Alaska Permanent Fund decreased the probability of an Alaskan child being obese by as much as 4.5% , and the dividends overall were associated with reductions in low birth weights .
During the U.S.’s basic income experiments in the 1970’s, receiving basic income was associated with improved infant health . When the US subsequently reformed its welfare programs to make cash transfers conditional upon work, infant mortality rates went up, in particular for foreign-born Mexican women .
Across basic income experiments in Seattle, Washington, and Denver, receiving basic income increased the number of adults pursuing continuing education, raised high school completion rates by 25 - 30% , and improved attendance, grades, and test scores, especially for children of low-income households . In North Carolina, children of low-income households receiving an unconditional $4,000 per year increased their educational attainment .
With basic income, the details are everything. It may cost anywhere from $219 billion to $3.4 trillion per year, depending on whether it’s universally given to all, or targeted towards those with lower incomes, whether parents get extra cash for children, and how much the payment actually is.
There isn’t enough research on the optimal payment amount. Neither is there enough research on the opportunity cost of larger basic income programs. While universal and targeted programs achieve similar outcomes, each carries unique positives and negatives.
The costs of smaller programs are relatively simple to offset, but larger programs will require new or higher taxes, raising challenges of political feasibility.
Basic income can cost anywhere from $219 billion - $4 trillion annually, depending on the details.
Negative income tax programs have a single cost - generally ranging from $219 billion to $1.09 trillion depending on benefit amounts, phaseout rates, and whether minors are included.
The relevant costs of universal basic income programs are two-fold, with both a gross cost (the total amount of money distributed) and a net cost (the total amount of money distributed, minus the clawbacks from higher taxes that fund the program).
The gross cost of a $12,000/year UBI, such as Andrew Yang's Freedom Dividend, is estimated to cost $2.8 trillion per year . The net cost of UBI proposals depends on the specific taxes given to fund the program, but may range anywhere from $539 billion to $1.69 trillion .
To benchmark against real proposals, here’s a sampling from across the spectrum:
The Wiederspan, Rhodes, & Shaefer Household Negative Income Tax
Goes to households, eliminates poverty (set at 100% of the poverty line) and phases out completely by 303% of the poverty line (33% phaseout rate).
Cost: $336 billion per year (for calendar year 2004, using 2007 dollars. $482 billion in 2021 terms).
Hartley & Garfinkel’s Small Guaranteed Income
$250 per month to all individuals 64 and younger. Begins phasing out once individual income surpasses $150,000.
Gross cost: $720 billion
The Kirwan Institute Guaranteed Income
Payment of $12,500 to all individuals earning $0, begins phasing out once earned income surpasses $10,000, and phases out entirely once single-adult households earn $50,000. Includes additional support of up to $4,500 per child.
Cost: $876 billion per year.
Andrew Yang's Freedom Dividend
$1,000 per month to all adults.
Gross cost: $2.8 trillion.
How to pay for basic income? Common financing options for smaller programs include: consolidating income-support welfare programs, closing tax credits and exemptions, and stronger tax enforcement. Larger programs will require additional taxation, from value-added taxes to carbon or land value taxes.
The cost of smaller basic income programs, such as Wiederspan, Rhodes, & Shaefer's negative income tax, can be covered by consolidating existing welfare programs, closing tax credits, or stronger tax enforcement.
In 2019, the government spent $361 billion on income support welfare programs ( $217 billion on the EITC, TANF, SNAP, and SSI alone ), most or all of which could be consolidated into any basic income proposal. Recent estimates suggest that the tax gap - the difference between taxes owed and taxes paid - totals roughly $600 billion annually . The Treasury Department estimates that an extra $80 billion in IRS funding over ten years could generate an extra $320 billion in tax revenue. Data sharing between the IRS and financial institutions is estimated to raise an extra $460 billion over the next decade.
Moderately more expensive programs, such as Hartley & Garfinkel's $250/month guaranteed income, can be mostly covered by closing existing tax credits. The additional cost may be financed via other changes, such as stronger tax enforcement, or as Hartley & Garfinkel propose, a carbon tax.
Larger programs, such as The Kirwan Institute's proposal, or Yang's Freedom Dividend, would require all of the above, plus larger, broad-based taxation. The most common being a value-added tax, which as of 2018, was in effect for all OECD countries except the U.S., and 166 of the 193 countries with full UN membership. A 10% value-added tax (half the average rate in Europe) is estimated to raise an average of $1 trillion per year .
Other taxes proposed to offset program costs include: carbon tax, land value tax, consumption tax, payroll tax, wealth tax, financial transaction tax.
One final funding source is to account for both the expected economic growth caused by basic income, and reductions in public expenditure, such as reduced rates of incarceration, healthcare expenditure, and welfare uptake. For example, Vancouver's cash transfer of $7,500 to recently homeless individuals more than made up the entire program cost by reducing the nights recipients spent in homeless shelters, reducing the necessary public funding.
In a similar vein, one estimate found that providing an unconditional child allowance that costs $100 billion per year could generate up to $865 billion per year in monetary and non-monetary benefits, again more than offsetting the program’s full cost. Reducing childhood poverty could also spur economic growth that offsets program costs in the long-run. Childhood poverty is estimated to cost the US $500 billion per year , reduce productivity and output by 1.3% of GDP , and raise the costs of crime by 1.3% of GDP and those of healthcare expenditure by 1.2% of GDP.
Basic income for all, or targeted to lower-incomes? Universal programs are simpler and more effective, but require more revenue. Targeted programs require greater administrative burden, perpetuate welfare stigma, but are more politically and economically palatable, largely due to requiring less funding.
Taking financing into account, universal and targeted programs can be, counterintuitively, identical in their distributional effects. Whether a program provides a universal transfer to all citizens financed by a flat tax, or a targeted transfer that gives a full amount to low income-citizens and phases out as earnings rise, the outcome is identical.
But this is not say there aren’t significant differences between these two approaches to a basic income.
Universal programs are easier to administer, more popular among citizens , and more highly correlated with higher levels of redistribution than means-targeted programs . But they require roughly double the level of funding to achieve similar outcomes , which some argue reflects a greater strain on society’s overall redistributive capacity .
Means-tested programs like NIT require fewer taxes to achieve identical redistributive effects. However, they require complicated administrative systems that match benefit payments to fluctuating incomes, create incentives to under-report income, create frictions that may keep the most vulnerable from receiving benefits, and perpetuate welfare stigma.
In a series of negative income tax experiments across the US in the 1970's, recipients consistently under-reported their incomes in order to receive higher benefit payments . A national NIT program would likely face similar reporting issues.
How much should basic income pay? There isn’t enough research on identifying the most effective amount, or weighing the opportunity costs.
Some advocates argue that a just society ought to provide the highest sustainable basic income to all citizens, while others argue that the payment should be just high enough to meet an individual’s basic needs. By unconditionally provisioning every citizen’s basic needs, the basic income could ensure that all participation in the labor market is truly voluntary , by virtue of maintaining a real alternative.
But how much is necessary to meet a citizen’s basic needs? Higher basic income proposals take the poverty line as a rough approximation. But any program large enough to provide a minimum of the poverty line to all citizens will carry significant opportunity costs.
Every dollar spent on basic income is one that could be spent elsewhere. There is likely a point after which increasing the basic income amount does less good than allocating funds to other kinds of programs. But as the Jain Family Institute writes, there exists insufficient research to identify this threshold :
Current research measures a hypothetical basic income against the status-quo of means-tested welfare programs like the EITC. There is little research that compares basic income against other hypotheticals, from job guarantees, expanded EITC programs, universal health insurance, to further investments in housing, education, or public services.
Basic income every year, or every month? Larger annual payments tend to go towards larger investments, while more frequent monthly payments reduce income volatility, and diminish debt. More research is needed.
The frequency of payments will likely affect patterns of consumption, savings, and well-being. Given that the ideal frequency may vary with individual circumstances, current research is exploring the feasibility of offering a choice to recipients .
In the context of the earned income tax credit, more frequent payments were found to reduce stress, diminish debt, and improve mental health .
Meanwhile, evidence from Kenya suggests larger, lump-sum payments tend to be spent on large assets, while more frequent payments lead to better balances of spending and saving .
Thanks for reading – we're working on developing easy ways for readers to suggest papers, insights, or experiments you think we've missed. Until then, if you think we should consider adding anything, let us know (thelibrary [at] economicpossibility.org).