[ SkipToMainMenu ]

A Research Agenda to Get More People Out of Poverty

This essay is part of Issues in Science and Technology’s “The Next 75 Years of Science Policy.” 

Today, far too many people live in poverty, even in high-income countries such as the United States and Canada. This situation hurts everyone. Poverty alleviation is not an intractable problem, but its politics have been unyielding. We argue that solutions can and must be found, especially because reducing poverty and inequality provides not just a greater sense of security, freedom, and dignity, but also, crucially, it unleashes human potential and creativity. In particular, research has shown that alleviating poverty increases cognitive bandwidth and frees up the intellectual power needed for productivity growth.

It has become increasingly clear that technological progress in free market economies will not eliminate poverty. It is equally doubtful that any one program of redistribution is equal to the challenge. But in principle the right mix of policies could allow for some redistribution without detriment to economic productivity. We propose here a significant governmental investment in research to test various promising combinations of programs to lift the most disadvantaged people out of poverty while maintaining a vibrant and innovative economy.

We believe such research is an important part of considering the future of science policy because eradicating poverty is itself an important social goal. In addition, many current science and technology policy proposals attempt to deal with poverty and inequality either directly or indirectly. Thus testing and evaluating policy packages that combine some of those proposals should be part of the public research portfolio.

We call this approach synergistic system design, which involves combining known social programs into policy packages that interact beneficially, so that their advantages reinforce, but not their disadvantages. We argue that this approach has the potential to significantly reduce poverty and that the required investment can be justified by the resultant societal benefits.


Even in high-income countries, poverty rates are disturbingly high. For example, using a metric called the Supplemental Poverty Measure, which adjusts the official poverty line by including other information, the Urban Institute projected that in 2021 one in seven Americans (13.7%) would live in poverty. In Canada, one in ten (10.1%) lived under the poverty line in 2019 (as defined by the Canadian government).

In our view, poverty in high-income countries is largely a problem of income distribution, not fiscal capacity. But it’s hard to redistribute income without affecting a nation’s productive capacity. The incentives and efficiencies necessary to generate a high gross domestic product (GDP) could be undermined by a poorly designed redistributive system that would keep all citizens out of poverty.

The US economy is a major example of the distribution problem. While its GDP per capita doubled over the last 50 years (in inflation-adjusted dollars), the official poverty rate remained unchanged. Arithmetically speaking, a small tax increase on that new wealth could have eliminated poverty entirely. Yet, political support for such a program is inconsistent and seemingly never enough to make it a national priority. Politicians have all but abandoned redistribution and have instead deposited their hope in technological innovation—reasoning that if innovation can deliver economic growth and constantly raise national income, all boats will rise with the tide. But the last 40 years have shown that more and more boats are sinking in no small measure because innovation tends to perpetuate and even exacerbate existing income inequalities. Importantly, we do not propose a return to the old recipes for redistribution; rather, we realize something new must be proposed, with the dual objectives of greater efficacy and improved political feasibility.


The synergistic system design that we propose can be illustrated with a simple example. Consider two ways local governments reduce automobile injuries: by encouraging seat belt use and by discouraging speeding. For either approach, a small enforcement effort provides a good return on investment, but with greater efforts, the rate of return diminishes. This means that it is better to share efforts between the two than to put all the eggs in one basket. Sharing efforts between different but related initiatives often yields a whole that is greater than the sum of the parts. That is the beauty of synergy.

However, synergistic solutions are not always easy to find because there are usually numerous possible permutations, many of which may not yield a net benefit. This is why we believe that a large program of antipoverty research should be deployed to find the most promising policy combinations and to design them to be as effective as possible. In addition to the necessary empirical research, mathematical modeling could help develop plausible hypotheses. There’s good reason for optimism because synergistic design has worked well in other fields. It led, for instance, to significant improvements in medical treatments with drug combinations that offer net benefits not possible with an increased dose of a single drug alone.

Various policies for poverty alleviation have been tried, but rarely in combination. Here are five examples of policies that could be part of a carefully crafted combination for achieving overall benefits that significantly exceed overall cost:

Minimum wage. A minimum wage boosts take-home pay for some employees, but it also has the potential to increase unemployment by making it unprofitable for employers to hire some workers. Many students of the minimum wage agree that it helps to mitigate poverty, but even they remain divided about its optimal level. What’s more, proponents of the minimum wage agree that, on its own, it cannot significantly reduce poverty.   

Basic income. This term refers to programs that provide each adult citizen with universal, unconditional regular cash payments to help cover basic living costs. Past and ongoing experiments have found wide-ranging benefits of cash transfers, including improving individual well-being and helping people start businesses, gain higher education, and take care of their health problems. Positive network effects observed included declines in crime and a migration away from other safety net programs.

Negative income tax. This policy is a way to increase take-home pay of the lowest earners, without forcing employers to pay a higher wage than they could afford for that employee in that situation. In a negative income tax experiment run in several US states between 1968 and 1982, workers below a certain income threshold received money. The results did show a small reduction in the labor supply, but the decrease was mostly driven by youth who opted out of work to attend school, which was probably beneficial overall.

Subsidized goods and services. In this approach, the government provides free or low-cost basic goods and services. Generally, this policy is popular for collective services that may be impractical to monetize, such as firefighting, policing, and national defense. However, poverty mitigation may require direct subsidies to meet basic needs such as food or energy.  

Wage subsidy. This approach is somewhat similar to the negative income tax, but in this case the government makes payments directly to employers to help subsidize employment costs. In turn, wage subsidies can encourage innovative organizations to create expanded employment opportunities while providing affordable new goods and services. A wage subsidy program could be made conditional upon the recipient firm’s achieving and maintaining measurable targets for safety, diversity, equity, and inclusion. Subsidy payments could also take the form of nonrefundable tax credits.

An example of synergistic design that should be tested is combining a basic income program for workers with a wage subsidy program for employers. In that approach, all citizens within a selected community would receive a basic income, with no required payback beyond regular income tax, and employers would receive a nonrefundable tax credit, making it affordable for them to hire employees with lower intrinsic productivity. The wage subsidy would encourage employers to offer a modest number of minimum wage positions to individuals who would not otherwise provide sufficient value to merit employment.


In the Manitoba Basic Annual Income Experiment that took place in Dauphin, Manitoba, in the 1970s, low-income families received an annual income guarantee between CAD 3,000 and 6,000 (roughly CAD 20,000 to 60,000 in today’s currency). The experiment resulted in a significant reduction in hospitalization, specifically for mental health diagnoses and work-related injuries. Recently, the New Leaf Project was conducted to examine the impact of unconditional cash payments on individuals experiencing homelessness. In the project, a one-time unconditional payment of CAD 7,500 (equivalent to personal annual income assistance in British Columbia) was made to each of 50 homeless individuals in Vancouver, Canada; another group of 65 served as controls. The cash transfer was in addition to existing welfare benefits. The results demonstrated that the cash transfers led to significant improvements in recipients’ standard of living, cognitive function, and subjective well-being. Notably, there were no increases in spending on alcohol and drugs. Somewhat surprisingly, the cash transfers actually produced net savings of CAD 625 per person per year because of reduced shelter use.

As a new pilot experiment in universal basic income, the city of Compton, California, launched the Compton Pledge in late 2020. The city provides up to USD 1,000 per month to qualifying families for two years, as unconditional, direct, and continuous cash transfers to supplement their existing welfare benefits. This pilot adds to a growing list of US cities (also including Stockton and Oakland, California) that are experimenting with basic income programs. 

Canada may be able to adopt a basic income program without adding to its fiscal debt. For example, the Parliamentary Budget Office of Canada has estimated that a guaranteed basic income of CAD 17,000 per individual (or CAD 24,000 for a household) would cost CAD 88 billion in 2022–23, and that this amount could be financed entirely by the removal of certain tax credits and some social assistance programs that overlap with the objectives of basic income. The simulation predicts a 50% reduction in the national poverty rate and a reduction in hours worked of less than 1.5%. Importantly, even that encouraging analysis overlooks the medium and longer-term benefits of poverty elimination that are not captured by a first-order economic analysis.


The proposed synergistic approach for solving poverty will require numerous experiments at a large scale in both size and duration. This experimentation is essential for producing reliable, replicable, and generalizable results. There is a growing awareness that use-inspired basic research, led by trans-sectoral teams, can be a powerful way to tackle difficult, complex problems. Those teams must include implementation experts and researchers who are pioneering new approaches to develop potent combinations of, in this case, poverty-reduction programs. Furthermore, since synergistic approaches blend numerous ingredients, a great many recipes could potentially be beneficial and therefore should be independently evaluated.

Yet another key consideration is to ensure that experimental antipoverty interventions operate for a long enough time to assess their impact on community culture. For example, it would be understandable to worry that a basic income policy, on its own, could gradually diminish the value that culture places on employment, even though short-term basic income experiments have not detected such an effect. This possible effect could be countered by another ingredient in the mix, as we suggested above. But that sort of dual-effect hypothesis needs to be carefully tested, at scale, over many years.

Overall, we need to better identify and understand three aspects of the issue: (1) the most cost-effective combinations of antipoverty interventions; (2) the optimal investments in these areas to achieve a satisfactory reduction of poverty; and (3) the means to communicate these discoveries in order to get buy-in from governments and the public.


Now is an ideal time to vigorously test this synergistic approach to alleviating poverty. The COVID-19 pandemic has forced many governments, especially in high-income countries, to experiment with economic and social policies such as stimulus checks, reforms to employment insurance to recognize gig work, and recognition of low-skill work as high-value activities, all while testing the capacity of the state to underwrite social programs. With this expanded understanding of what is possible, we argue that the time has come to do the required research to develop practical synergistic solutions. Put simply, the risks associated with not doing this essential research would vastly outweigh the comparatively modest costs of advancing this significant research agenda.

We therefore call for major investment, with a consequent obligation for excellent planning, to devise the optimal blend of interventions—first for testing and then for implementation. There is no doubt that high-income countries have the resources to accommodate such an investment.

Consequently, we believe it is time for a massive, Apollo-scale research investment that combines basic and applied studies of large-scale interventions to establish better paths forward for overcoming persistent poverty.