Are we wrong about Redlining?

In poverty and inequality circles, redlining is often one of the first things mentioned when talking about why racially-correlated urban inequality exists today. When I previously wrote a blog post about mapping poverty in my hometown of Saint Paul, I mentioned the impact of redlining on current patterns of inequality in the city. 

Redlining as most people know it refers to a series of maps produced by the Home Owners’ Loan Corporation in the 1930’s, where neighborhoods were graded by their social, economic, and physical characteristics. The neighborhoods marked in red on these maps line up with neighborhoods that today have low incomes. 

However, new research on these maps shows that this commonly-held belief about these maps might be wrong. 

Before digging in any further, I want to make it clear that this research does not suggest that there isn’t historical discrimination that led to the spatial trends we see in urban areas. Instead, this research tries to show that the Home Owners’ Loan Corporation maps are not the main culprit. 

What are the Home Owners’ Loan Corporation maps?

During the Great Depression, the Home Owners’ Loan Corporation was established as part of the New Deal in order to help prevent a total housing market collapse. Their goal was to purchase failing mortgages from banks and restructure them so people could stay in their homes.

The Residential Security Maps and their color-coded neighborhoods were created in an attempt to measure the riskiness of neighborhoods instead of individual borrowers. To determine the riskiness of a neighborhood, the evaluators looked at features like housing quality, access to public transportation, economic characteristics of the people who lived there, and the race of the people who lived there.

It is true that almost all majority black neighborhoods in the cities evaluated by the Home Owners’ Loan Corporation were given the lowest grade, but it is important to remember that in the 1930s, discriminatory housing practices largely kept people of color out of urban areas. The majority of the lowest-rated neighborhoods were almost exclusively white immigrant neighborhoods with poor housing conditions. 

In the predominantly black neighborhoods during those years, housing conditions were even worse. This means that even if those neighborhoods were not predominantly black, they would have still received the lowest rating because of their housing conditions. 

In fact, during the time that the Home Owners’ Loan Corporation was purchasing mortgages, they did a fairly admirable job of providing assistance indiscriminately. This is not to say that they were actively supporting disadvantaged communities, but they purchased a proportionate amount of loans from the lowest rated neighborhoods.

This begs the question though, if these maps were not made with discriminatory housing practices in mind, what were they for?

Unfortunately, there isn’t a satisfying answer to this question. By the time these maps were completed, the Home Owners’ Loan Corporation was largely out of the mortgage business. The maps weren’t shared with private lenders, and although they were shared with the Federal Housing Administration (which did participate in discriminatory housing practices), that agency had its own set of maps, and the available historical evidence suggests that they didn’t use the Home Owners’ Loan Corporation maps. 

Why are these maps correlated with modern disadvantaged communities?

The history lesson above is fascinating, but it doesn’t explain why the Home Owners’ Loan Corporation maps are strongly correlated with modern day low-income neighborhoods. If they weren’t used to perpetuate discrimination, how did we get here?

The answer lies in understanding what the maps were actually measuring: housing quality. 

Remember, when the maps were made, predominantly black neighborhoods were rare in America’s cities. Racist housing policies largely kept people of color out of the cities altogether. 

This changed after World War II when we saw one of the most dramatic realignments of people in the United States. The simultaneous impacts of the Second Great Migration and White Flight completely changed the demographics of American urban environments. 

The discriminatory housing practices that kept people of color out of the cities vanished and were replaced with new discriminatory housing practices that kept them out of the newly built suburbs. White families from the cities started leaving en masse, leaving vacancies for black households. 

Before this migration, black urban neighborhoods had a lot of economic diversity. It largely didn’t matter how much money black households had, the racist housing policies made sure they all lived in the same neighborhoods. 

During the migration, black households had a new opportunity to choose where they lived. Black urban neighborhoods became economically separate, and wealthy black households had the opportunity to move into the higher-rated neighborhoods.

Another key factor was that in the lowest-rated neighborhoods, far more families left for the suburbs than moved in to replace them. This led to housing abandonment and disinvestment in these neighborhoods that helped perpetuate the economic discrimination. 

So, the reason these maps line up with racial and economic disparities today is because during this massive upheaval of American cities, poor black families moved into neighborhoods with poor housing conditions that were vacated by white families who moved elsewhere. The housing markets in these neighborhoods largely collapsed due to high vacancy rates caused by fewer families moving in than previously lived there, which coupled with racist economic development policies led to disinvestments in these neighborhoods.

What does this mean for policy today?

If you believe everything I’ve said up to this point, you might still wonder why this distinction matters. At the end of the day, the Home Owners’ Loan Corporation maps line up with disadvantaged neighborhoods today, and isn’t that really what matters?

While I agree that addressing today’s problems is the most important issue, understanding how we got here can help us come up with permanent solutions. 

One takeaway I get from this new understanding is that the low-income neighborhoods in our cities today were not the result of black people being forced into those places by federal agencies. They were forced into those neighborhoods because of their wallets. 

We know this because wealthy black households moved into higher rated neighborhoods during the Second Great Migration. The households that moved into these lower rated neighborhoods did so because that was what they could afford to do. 

Another reason it might be good to not allow these maps to influence policy today is because while the low rated neighborhoods do largely line up with low-income neighborhoods today, it’s not a perfect match. If policy were designed to only focus on the lowest rated neighborhoods from the 1930s, we’d be ignoring a lot of disadvantaged communities that don’t fall exactly in those neat lines.

Finally, focusing too much on these maps makes it so we do not accurately acknowledge the players that were responsible for racist housing policies in that era. There were bad actors who made it extremely difficult for black families to find adequate housing, and we should accurately tell that part of our history. 

I would encourage anyone who has read this far to go and read the full article that helped me understand this. There is much more to this story and Alan Mallach does an excellent job communicating it. It’s always an unusual experience to have an idea that you have accepted as a fact be challenged this way. It took me a while to come around to the idea that I had been wrong about what these maps meant.

Why do we keep subsidizing big homes for wealthy families?

This month, researchers Michael P. Keane of Johns Hopkins University and Xiangling Liu of Hunan University released a paper in the National Bureau of Economic Research’s working paper series on reforming taxation of housing.

This working paper revolves around a problem in housing economics called imputed rent.

Say you rent your home. That means you pay a landlord for the ability to live in your home. Since you are paying someone, they receive that payment as income. Like any other income, this income is subject to income taxes.

What if you live in your home? Well, you pay your mortgage, the same way a landlord does. But you don’t pay a landlord for rent. You “pay” yourself to live in the home that you own. You get the same good (housing) as any renter gets, but since you own the home, you pay yourself from it. But here’s the catch–that payment is tax free. Because no money changes hands, you are exempt from income taxes.

What this amounts to is an implicit subsidy to own a home rather than rent. This theoretically pushes people who would be renting to own due to the subsidy, in particular encouraging upper-income households to buy larger houses than they would otherwise. It also ties up capital in homeownership. Rather than investing in companies, individuals spend their incomes on homes, artificially inflating their value and decreasing the value of businesses.

These researchers investigated what would happen if an alternative tax system existed. Under this system, all homeowners would be treated as landlords and people who lived in their own homes would be required to report the imputed rent they paid and received by living in their own home as income. Just like landlords, homeowners would be allowed to deduct property taxes, maintenance, depreciation, and mortgage interest. After all this was deducted, however, they would have to pay taxes on the value of rent they are paying themselves.

This would mean a lot less money in the pockets of homeowners and a lot more money in the pockets of the government, right? Not necessarily. Keane and Liu model what would happen if this extra government revenue was used to finance an across-the-board income tax cut for all taxpayers. This tax cut would offset part of the new tax for homeowners.

…or rather, all of the new tax for homeowners. Keane and Liu simulate this policy change using data from the Panel Study of Income Dynamics. They estimate that this new tax would finance a 9.15% income tax cut across the board. This means every 11th dollar raised by income taxes would go back to households. It also leads to a rare policy change that would lead to a Pareto improvement, meaning no one will be left worse off by the policy and some (many in this case) will be left better off. Under their simulation, every homeowner paying taxes on their imputed rent will have their new taxes offset by the tax cut and the average household will see incomes raised by 0.79% after taxes. This means the average household will get an additional $610 every year under this policy.

Furthermore, housing prices would fall modestly under this policy due to a moderation in demand for housing. Keane and Liu estimate this would mean a 0.7% reduction in average housing price, which would amount to about $2,400 for the average U.S. home.

An interesting wrinkle to Keane and Liu’s simulation is that they actually don’t find that the homeownership rate would decrease, which on its face seems counterintuitive. Why would it depress prices on homes but not the homeownership rate?

Their simulation finds the answer for this: people would reduce the size of their homes. The current uneven playing field between owning and renting doesn’t actually get more people to own homes. It mainly incentivizes ownership of larger homes.

The researchers also use their model to estimate the impact of two proposals to eliminate the mortgage interest deduction.

How the mortgage interest deduction works is that a household can subtract the amount they pay in mortgage interest from their reported income. So this means a wealthy household paying $24,000 in interest and $9,000 in property taxes could effectively save $12,000 from the mortgage interest deduction.

The mortgage interest deduction is an unpopular policy among economists because it essentially functions as a subsidy to high-income households. Since low- and middle-income homeowners usually take the standard deduction, they do not benefit from this deduction. So this means the more wealthy you are and the larger your house is, the more you benefit from the mortgage interest deduction.

Keane and Liu estimate what would happen if the mortgage interest deduction were eliminated and replaced with an across-the-board income tax cut. According to their simulations, replacing the mortgage interest deduction with an income tax cut would drop income tax rates by 4.7%, housing prices by 1.66% (about $5,600 in 2023), and would discourage homeownership slightly, dropping homeownership rates from 64.9% to 64.3%. Incomes would rise by 0.76% (About $590 for the average household) and it would also be a Pareto improvement: all households would have more income under this policy. Benefits for this change would be highest for low- and middle-income households and higher-income households would purchase smaller homes under this scenario.

They also put forth an alternative: replacing the mortgage interest deduction with a revenue-neutral refundable tax credit. This means all homeowners could claim the credit and upper-income homeowners would not have higher benefits than lower-income homeowners. Under this scenario, ownership rates increase from 64.9% to 68.7%, home prices fall by 1.3% (about $4,400), average income rises by 0.58% (about $450 for an average household), and low- and middle-income households gain the most. High-income households end up worse off under this scenario due to not benefiting from the income tax cut in the other mortgage interest deduction elimination.

A lot of money is tied up in subsidies to high-income households to pay for larger homes. Eliminating these subsidies can lead to scenarios that make everyone better off, free up investment for more productive uses, and help low- and middle-income households. We’ll see if Congress pays attention to this well-known problem during its tax reform work this year.

Rent Eats First: The Problem With Our Housing Cost Burden

In September 2024, the U.S. Census Bureau published a news release titled “Nearly Half of Renter Households Are Cost-Burdened, Proportions Differ by Race.” Their analysis found that 49.7% of all American renters are spending at least 30% of their income on rent—a threshold commonly used to define “housing cost burden.”

The report also found that 56.2% of Black or African American renter households exceed this 30% threshold, representing about 4.6 million people. For many of these households, rent is the largest and most inflexible monthly expense—leaving less room for food, healthcare, childcare, transportation, or savings. In short: rent often eats first.

But where does this 30% benchmark come from? Who decided that this ratio was a reasonable measure of affordability?

The answer lies in a set of policy decisions that began in the mid-20th century. In the 1960s, economist Mollie Orshansky at the Social Security Administration developed the federal poverty thresholds based on Department of Agriculture food budgets. At the time, American families spent about one-third of their income on food, so Orshansky multiplied the minimum food budget by three to estimate what a “poverty-level” income should be. Around the same time, U.S. housing policy began adopting a similar one-third rule to judge how much a household could afford to spend on housing. The 30% standard was formally codified in federal housing programs in the early 1980s.

While the 30% threshold is still widely used today by HUD and other agencies, critics note that it may no longer reflect the financial realities of modern households. The cost of housing, transportation, and healthcare has far outpaced income growth in many regions, and a flat percentage may not capture the full picture of affordability, especially for low-income households.

For example, in 2025, the federal poverty line for a family of three is $26,650, or about $2,220 per month. With the national median rent at $1,406, that family would be left with just over $800 to cover all other expenses. In such cases, even spending less than 30% of income on housing can result in financial strain.

So what can be done?

Instead of using a fixed 30% calculation, policymakers could evaluate whether households have enough money left over to cover expenses after rent has been paid. This is called “Residual Income Measures,” and it’s a different way of configuring a budget. The system we have now plans on rent eating first. We look at someone’s monthly income and divide that by three to figure an acceptable monthly housing cost. Residual Income Measures looks at what is left over after all other basic necessities have been met. 

Additionally, we could localize housing cost burden. For example, rent in San Francisco, California is very different from rent in Little Rock, Arkansas. Because of the difference in localities, affordable housing should be considered within its regional context. Two individuals each making $60,000 in San Francisco and in Little Rock will have very different socioeconomic situations. Taking into account regional differences in calculating housing cost burden could improve our calculations. 

Policy solutions for addressing housing cost burden are a different conversation entirely. For example, expanding Section 8 Housing Vouchers could help families who make too much money to be considered “in poverty,” but still navigate daily life with incomes near the poverty line without government assistance. Right now, waitlists are long. Expanding eligibility for housing assistance can fill in some of the gap that exists for families who make too much to qualify but still struggle to cover monthly expenses. 

Additionally, lawmakers could incentivize affordable housing development. Increasing the supply of affordable housing can help meet the growing demand for affordable housing. This can be accomplished by zoning reforms and through nonprofit partnerships. Nonprofits often work directly with individuals most afflicted by high housing costs, and can directly connect the people in need to the new affordable housing developments. 

In the end, long-term solutions may benefit from integrating housing policy with broader anti-poverty programs. Housing cost burden doesn’t occur in a vacuum. Coupling housing support with wraparound case management has the potential to achieve long-term gains. This could look like a single application or website where an individual can apply for Medicaid, SNAP, TANF, and Section 8 vouchers.

What does universal pre kindergarten do for parents?

Last month, we published a cost-benefit analysis of the impacts of a potential universal pre kindergarten program in Ohio. We estimated that a universal pre kindergarten program would lead to between $220 and $750 million of benefits for Ohio, largely in the form of higher future earnings for the children who would be enrolled. 

When we did this analysis, we focused our analysis on impacts that would be realized down the road. Benefits such as future earnings, reduced future crime, and lower participation in special education programs are all beneficial, but they don’t help people today.

When we look at future benefits, we always make sure to discount them so that we capture how much people are willing to pay today for those benefits in the future. There is a lot of debate about how exactly to do this, but it's important to remember that when we are conducting a cost-benefit analysis, we are trying to capture how much people today care about investing in the future. 

A big part of the reason we care about benefits that don’t get realized until far into the future is because there has been a growth of evidence of these long-term benefits in economic research. Recent advances in the study of intergenerational impacts led by the legendary economist James Heckman have shown how the benefits of providing assistance to very young children can compound over time and create huge benefits for years to come. 

However, new research shows that universal pre kindergarten may not just be a long-term policy, it can deliver meaningful short-term economic benefits as well.

A new working paper looks at the effects of universal prekindergarten programs across nine states, specifically focusing on the short term employment impacts. The findings show that universal prekindergarten programs not only increased enrollment in early education, but also boosted labor force participation, employment, and hours worked, especially among mothers of young children. These programs reduced child care constraints, enabling more parents to work or work more hours.

Interestingly, the paper found that the benefits weren’t just limited to parents of young children. Other women, such as informal caregivers or those considering starting families, also saw employment gains, demonstrating the fact that universal pre kindergarten creates positive externalities. 

Importantly, the size of these effects varied by location. There were larger economic gains in areas with higher enrollment and stronger program quality. In other words, well-designed and well-attended universal pre kindergarten programs don’t just help children, they function as a form of economic stimulus, boosting household earnings and labor force participation in the near term.

The authors of this paper also calculated how much additional tax revenue would come as a result of these employment gains. Amazingly, they estimate that there might be enough additional tax revenue to cover the upfront costs of the program entirely. This fiscal impact means that universal prekindergarten might not represent much of a short term burden at all, since it would not have to take away money from other short term public programs.

One limitation of this study is that it did not value time spent at home, effectively ignoring the major cost of the program. While labor market time is valuable for employers and employees, it requires parents to give up nonmarket time, which is valuable on its own.

So, while the long-term benefits of universal pre kindergarten are substantial and well-documented, this new research reminds us that the economic case for universal prekindergarten isn't just about the future, it's also about helping families and communities today.

How do cigarette taxes impact household budgets?

When Governor DeWine announced his budget proposal earlier this year, there were a lot of major changes that were fascinating for policy analysts. At Scioto Analysis, we’ve written about things like stadium subsidies, library funding, and especially the Child Tax Credit.

Another important part of the Governor’s budget was an increase in taxes on cigarettes, marijuana, and gambling, collectively referred to as “sin taxes.” My colleague Rob Moore wrote about the benefits of taxing these goods from an economic perspective, but he also acknowledged the most frequently cited downside: sin taxes are regressive. On average, lower-income people spend a larger percentage of their income on these goods, and as a result they bear most of the burden of these increased taxes. 

When faced with higher prices, people will usually consume less of a good. From a theoretical perspective, we often ignore the actual mechanism of people participating less in a market. The effect of one out of 100 people quitting smoking is the same as all smokers reducing their smoking by 1%. 

In practice though, people have different responses to increased prices. Some people who have a low willingness to pay for these goods will stop consuming them altogether, people with a moderate willingness to pay may adjust their consumption by a little, and people with a high willingness to pay will not change their consumption and instead just manage the higher prices. 

The third group is the one I want to highlight today. 

A new working paper released this month looks into the question of how households change their consumption habits when faced with higher cigarette taxes. They use two approaches to answer this question. First, they surveyed current smokers and asked them how they would respond to a hypothetical price increase. Then they looked at actual consumption data and quantitatively assessed how people responded when taxes went up. 

The survey responses aligned with what we expect to happen in practice. Some people said that if faced with higher prices they would try to quit, some people said they would try to reduce their smoking habits, and some people said they would just deal with the higher prices. 

The most interesting findings came from the quantitative analysis. When households (particularly low-income households) are faced with higher cigarette prices, they tend to reallocate their spending away from what the authors describe as “human capital forming expenditures.” These authors suggest that households offset nearly 70% of the increased cost of cigarettes with reduced human capital expenditures. 

Previous research on cigarette taxes has shown that higher taxes lead to better human capital outcomes, such as better health and higher education, despite lower spending on these goods by households. This seems to suggest that the benefits of overall reduced consumption of cigarettes outweighs the increased cost and resulting reduction in human capital investment. 

One commonly suggested policy option for reducing the regressivity of taxes is to take the additional revenue and rebate it back to the households that bear the burden. In this case, the state could use the revenue generated from a tobacco tax and use it to subsidize the human capital-developing goods that people consume less of as a result. Think financing college scholarships for low-income households with tobacco tax revenue.

Regressive sales taxes present challenges for policymakers and families. This paper highlights undesirable consequences that come from increasing the costs of goods that have negative externalities. However, benefits still may outweigh the costs depending on the exact structure of the tax, and with some careful planning many of the downsides can be offset.

What is the difference between economic impact analysis and cost-benefit analysis?

“New study finds cycling has $1.4 billion economic impact on Iowa each year.” 

This was the headline for an article published in the Des Moines Register in January of this year. It was covering a study we did on spending by cyclists in Iowa.

“Study: Adult-Use Cannabis Legalization in Ohio Would Generate $260M in ‘Net Benefits for Society’”

This was the headline for an article in cannabis industry news website Ganjapreneur in 2023 on a study we did on a ballot initiative to legalize recreational marijuana in Ohio.

While these headlines seem deceptively similar, we used very different methodologies to come to these two numbers. Ultimately, these two statistics are referring to two different things entirely.

The former article was an approach called “economic impact analysis.” Economic impact analysis works by finding a base spending figure then estimating how that spending ripples throughout the economy. We do this using a concept called “economic multipliers.” 

While it sounds technical, the intuition behind multipliers is simple. When you spend money at a bicycle store for example, the business spends that money on more bicycles and wages for its employees. Employees then spend that money on rent, groceries, transportation to work, and the bicycle wholesaler spends that money on buying bicycles from manufacturers and paying its employees. Then the bicycle manufacturer spends that money on employees and parts, and so on.

Researchers at the Bureau of Economic Analysis have analyzed spending patterns across the country and have used that data to estimate how much a dollar spent in one industry spurs new spending in other industries. We can use this to estimate how much spending in one industry ends up impacting other industries, how it grows wages, and how many jobs are supported by an industry.

The latter study on cannabis legalization works differently. For this study, we utilized a framework called “cost-benefit analysis.” Cost-benefit analysis concern more spending patterns, instead looking at the economy in a broad sense by including nonmarket impacts. Because analysts conducting cost-benefit analysis monetize outcomes that are not traded on markets, they do not have precise data to work with that they do when conducting an economic impact analysis. Often, cost-benefit analysis involves using studies on the impacts of policies in certain places and making adjustments based on local conditions to estimate where the impacts of a policy could be elsewhere.

Early childhood education is a great example of this. Most of the work done by Nobel Laureate James Heckman and Leading Economic Development Economist Timothy Bartik relies on studies such as the Perry Preschool Project and the Abecedarian Project. These experiments were randomized controlled trials conducted in the 1960s and 1970s with follow-up surveys collected in the decades that followed. These studies allowed researchers to estimate the long-term impacts of programs by comparing a treatment group to a control group that did not participate in the early childhood education program. We assume that the impacts of these programs in the 1960s and 1970s through today will be similar to the impacts of programs implemented today 50 and 60 years into the future.

This isn’t to say that economic impact analysis does not come with its own assumptions. Multipliers an analyst uses today are based on economic conditions a few years old. Upheavals like the COVID-19 pandemic, Brexit, or new tariffs across the economy can reshape the economic landscape in a short amount of time, making data collected just a few years ago less informative to analysis done today. But they still provide a starting place for analysts and decision makers.

One thing economic impact analysis has come under fire for over the past decade is its use in promoting subsidies for stadiums.

Often sports teams justify public subsidies with economic impact analysis studies showing hundreds of millions of dollars of economic impact associated with the investments. These are often the result of restaurants and bars developing around a sports stadium after construction, generating new economic activity where there was no activity before.

The reason economists are skeptical of these studies is the lack of accounting for opportunity costs. Economic impact analysis, by its design, only tells an analyst what the benefits of economic activity or a project are. They do not address costs.

The main costs economists worry about around stadiums is the entertainment that they displace. Since people are spending money at restaurants and bars in close proximity to a stadium, they are not spending their money elsewhere in the city. Since entertainment budgets are not very flexible (people do not suddenly stop spending money on rent, groceries, and gas to go to a Browns game because of a new stadium), new developments usually means displacement of economic activity elsewhere.

This does not address other costs, such as the tax drag created by financing the subsidies and potential social costs such as violence associated with bars in proximity to the stadium.

This is where cost-benefit analysis can be a more useful tool for analyzing the impact of a project. A well-designed cost-benefit analysis will give a policymaker an understanding of the major economic impacts of a policy–positive and negative.

This does not mean a cost-benefit analysis is superior to an economic impact analysis. They answer different questions. An economic impact analysis answers the question of “how big this project or sector is in the grand scheme of the economy.” A cost-benefit analysis answers “what the net impact adoption of a policy or program will have on the economy.”

Economic impact analysis can confuse the public if not explained clearly. Even with all the work we have done to explain the limitations of economic impact analysis and what exactly it tells us, reporters are human. They still make simple mistakes like comparing spending on a program to its economic impact and trying to make back-of-the envelope return on investment calculations. What you need to keep in mind is this: economic-impact analysis helps you understand how big an impact a sector or policy is, not whether its benefits outweigh its costs.

The long-term effects of abandoning Ohio’s Fair School Funding plan

Early this budget season, Ohio House Speaker Matt Huffman said school funding cuts were on the horizon. His House budget certainly follows up with that promise.

According to analysis by the Ohio River Valley Institute, the FY2027 school allocation under the House Plan falls $2.7 billion short of what the General Assembly agreed to invest in public education in the 2022 Fair School Funding Plan. That represents a funding cut of about 25% from the previous plan.

These cuts will be felt across the state. According to the same analysis, 91% of school districts will have less funding under the House plan than the Fair School Funding Plan. An unlucky 26 school districts will see their state support reduced by 50% or more.

In February, my firm Scioto Analysis asked 17 Ohio economists for their thoughts about the plan to cut spending on public education. Of those economists, 14 agreed the cuts would hurt Ohio’s economy in the long run. Only one disagreed.

Dr. Kathryn Wilson of Kent State University explained the harms reductions in school spending can have on the economy, saying they can lead to lower human capital development that hurts the productivity of future workers, but also that it can lead to more costs for taxpayers with more government assistance and criminal justice spending needed with a less educated state population.

In 2023, we conducted a cost-benefit analysis of school spending in Ohio. We built off evidence of the relationship between school spending, test scores, and graduation rates to estimate the long-term impacts of school spending on labor force productivity. We found that increased investment in students leads to wage impacts in the long run that will grow the state’s economy. We also found cuts will hurt productivity and reduce output for the state.

According to the Ohio Department of Education and Workforce, Ohio has about 1.7 million children currently enrolled from Kindergarten to Grade 12. This means the proposed $2.7 billion cut would represent about a $1,600 per-student decrease in spending from the baseline of the General Assembly’s Fair School Funding Plan.

In our 2023 study, we estimated what would happen if the state reduced school funding levels to the per-student expenditure in Indiana, which is about $3,600 lower than Ohio. We estimated this would cost the state somewhere between $30 billion and $120 billion in economic value in the long run.

Scaling these losses to match the Ohio House’s $1,600 reduction in per-pupil spending, we can estimate the reduction in statewide school funding will cost the state economy somewhere from $14 billion to $54 billion in the long run in the form of lower earnings from lower test scores, lower graduation rates, and higher social spending.

Yes, $2.7 billion is a lot of money. But educating a state workforce costs money. Cutting corners on education might lead to short-term benefits, but there are long-term costs the state will have to bear for decisions like this. These include lower productivity, lower earnings, and higher spending on social services and criminal justice.

This commentary first appeared in the Ohio Capital Journal.

Do school vouchers help students?

Last year, my colleague Rob wrote a piece for the Ohio Capital Journal about Ohio’s private school voucher program. This is the system by which families can get public money to help pay for private school. Essentially it allows qualifying families to choose a private school over whatever their local public school would be. 

Proponents of school vouchers argue that this allows families to choose whatever education works best for their children. Opponents argue that these vouchers shift public funds from public schools to private schools, worsening conditions for those who remain in public schools. 

One question that largely goes unanswered in any debates on this topic is what private school vouchers actually do for students. 

It’s not enough to just compare private school outcomes to public school outcomes, that ignores the self-selecting nature of private schools. We also should be interested in how outcomes change for people who don’t receive these vouchers, since their education situation is changing as well. 

Thanks to a new paper from researchers at the Urban Institute, we can now answer these questions. 

One important caveat to note is that this research focuses on Ohio’s school voucher program when eligibility was restricted to low income students. It has since seen a significant expansion, with nearly all Ohio families now qualifying for at least some benefit.  

This research found that students who enrolled in the school voucher program attended and graduated from college at higher rates than their peers in public schools. These impacts were felt most strongly by black students and students from the lowest income families. 

Perhaps an even more encouraging finding was that students who were eligible for school vouchers but chose instead to remain in public school also saw modest increases in their rates of college enrollment and graduation. It appears that by allowing these families to make a more flexible decision about their education creates positive externalities for other students too.

I will admit, I was pretty skeptical about school vouchers before reading this paper. I was particularly hesitant about what this program might mean for people who don’t participate in it, but the evidence speaks for itself. 

However, I don’t think this paper means that school vouchers are always good no matter what. I’ve written in the past about some of the challenges that arise when policymakers try to scale up pilot programs. Just because this program is effective for one group does not mean it will be effective for everyone. 

I’m curious to see how many students change schools as a result of this expansion. My prior assumption is that families with higher incomes who are now exposed to this subsidy might already have the requisite income to choose private school if they desire. If that is true, then this subsidy won’t change people’s education behaviors, but instead will just free up income for other types of consumption. However, if the expansion does manage to change people’s education choices, I now suspect that it will create positive outcomes for students. 

Free School Lunch: How we got it and where it’s going

The USDA reports that in 2024, 20.5 million children received free lunches and 900,000 children received reduced-price lunches. While the square pizza in the American Public Education system seems omnipresent, it’s a relatively recent phenomenon. 

School lunches were born out of the postwar era, when President Truman signed the National School Lunch Act of 1946. The policy aimed to stabilize the agricultural labor force and reduce chronic child malnutrition. The number of people engaged in farming dropped 67% in 20 years, from 17 percent of the total workforce in 1940 to six percent in 1960. The National School Lunch Act supported farmers by creating a guaranteed market for agricultural goods. In addition, policymakers saw school lunches as a way to address child malnutrition.

The school lunch program has been a convenient partnership for agricultural surpluses, especially dairy. As post-war agricultural production ramped up, the federal government often found itself with too much milk, cheese, and butter on hand. To prevent these goods from going to waste and allowing prices to tank, the USDA bought the excess and distributed it through schools. 

By the 1980s, this relationship with the dairy industry became so significant that critics referred to government stores as “the cheese caves” due to the massive stockpiles of processed cheese. School lunches became a way to stabilize agricultural markets while feeding children, making items like milk and cheese fixtures in cafeterias nationwide. 

While school lunches were federally supported by the mid-20th century, school breakfasts were not institutionalized until later. The National School Lunch Program expanded steadily after its initial rollout. The Child Nutrition Act passed in 1966 which laid the groundwork for school breakfasts. Support for these programs, however, often reflected the political will of the moment. Grassroots organizations filled the gaps of the school meal program.

One of the earliest large-scale, community-led efforts to solve child hunger came from the Black Panther Party. Huey P. Newton and Bobby Seale founded the Black Panther Party in 1966 and they served their first free school breakfast in January of 1969 within an Episcopal church. Without government funding, Panther members solicited local grocery stores for donations, consulted with nutritionists to determine what would make a good breakfast, and then got to work serving it up. The children received chocolate milk, eggs, meat, cereal, and fresh oranges.

School officials immediately reported results in kids who had free breakfast before school. “The school principal came down and told us how different the children were,” Ruth Beckford, a parishioner who helped with the program, said later. “They weren’t falling asleep in class, they weren’t crying with stomach cramps.”

In 2010, President Obama signed the Healthy, Hunger-Free Kids Act. This policy introduced more nutritious food into school lunches and aimed to reduce the number of children diagnosed with obesity within one generation. A study conducted on children who ate the reimbursed meals before and after the introduction of the policy found that their Healthy Eating Index scores increased by 30% for low-income students, 31% for low- to middle-income students, and 19% for middle- to high-income students.

In more recent years, attention has turned to the impact of school meals on students’ academic performance. In 2017, the Brookings Institution published research on how the quality of school lunches affects test scores. Among students enrolled in 9,700 California schools, access to healthier school lunches was associated with improvements in test scores of 0.03 to 0.04 standard deviations, or roughly four percentiles.

Reducing class size is another strategy aimed at improving test scores. The same Brookings study compared the relative cost of healthier school lunches—around $80 per student per year—to the gains achieved through smaller class sizes. Reducing class size by one-third cost about $2,000 per student in 1999. While reducing class sizes requires hiring more teachers and thus comes with higher labor costs, this approach is about five times more expensive than improving school lunch quality for a comparable increase in test performance.

In response to the instability wrought by the COVID-19 pandemic, the federal government picked up the tab for universal free school lunches provided to every public school student in the country. For most states, this policy expired in 2022, revealing the unmet need of many school children. Six states—California, Connecticut, Maine, Massachusetts, Nevada, and Vermont—opted to bolster their school lunch programs with state dollars immediately following the federal withdrawal of support in 2022. As a result, food insufficiency among school-aged children was 1.5 percent higher in states that did not extend universal free meals into the 2022–2023 school year compared to those that did. 

What began as a way to deal with agricultural surpluses and reduce child hunger has become a significant social safety net element for American children. The presence of milk and cheese in American lunches is more than a dietary choice, it's the legacy of economic policy. With more than 20 million children reliant on the school lunch program, Harry Truman’s words still ring clear: "In the long view, no nation is any healthier than its children or more prosperous than its farmers."

Can being a better neighbor save lives?

A few weeks ago, I wrote about new data from the World Happiness Report on the impacts of sharing meals. I encourage everyone to go read the full report in that post, but in case you haven’t already done that I wanted to talk about another finding I found interesting. 

Chapter Six of the report focuses on the connection between prosocial behavior and deaths of despair. For context, “prosocial behaviors” are activities like volunteering, donating, and offering help to strangers. “Deaths of despair” is a term for deaths due to suicide, alcohol abuse, and drug overdose. 

In the United States, we have a particular problem with deaths of despair. Between 2000 and 2019, the United States had the largest increase in the rate of deaths of despair among countries in the World Happiness Report. Among the countries that the World Happiness Report has data for, the United States does not have the highest rate of deaths of despair, but most other countries saw decreases in their rates over this same time period. Even if we include the rise in deaths of despair in the United States, the average rate across the globe has decreased. 

The new research highlighted in the report finds a connection between the rates of prosocial behavior and the rate of deaths of despair. Their regression analysis estimated that for every 10 percentage point increase in the share of people participating in prosocial activities, there is one fewer death of despair per 100,000 people. People age 60+ benefit even more, with an effect size nearly double compared to the general population. 

Those numbers are a bit hard to understand, so let's put it into context. In high income countries like the United States, about 35% of people participate in prosocial behaviors. If we increased that by 10 percentage points to 45%, that would mean an extra 34 million people engaging in prosocial activities. That would result in one fewer death of despair per 100,000 people, or about 3,400 per year across the country. 

Simplifying the math a bit further, for every 10,000 additional people who participate in some type of prosocial activity, one fewer person will suffer a death of despair each year. That seems like a pretty achievable goal to me. 

From an economic perspective, encouraging people to participate in prosocial behaviors seems like a very efficient way to decrease deaths of despair. Helping a stranger is often a very low cost activity, sometimes only taking seconds out of your day. 

Unfortunately, this is not necessarily something that policymakers have a lot of influence over. This is the same issue with trying to encourage people to share meals with each other: public policy is not very effective at changing social behaviors. 

Still, there are some things policymakers could explore to help encourage prosocial behavior. We’ve done past research that shows similar interventions are beneficial to the public. There also already exist tax incentives for people to donate money. Maybe there could be a way to incentivize people to volunteer their time in a similar way, like the Illinois volunteer emergency worker credit. 

Hopefully this information can encourage more people to go out of their way to help people in their communities. I know that I probably don’t engage in enough prosocial behavior myself, and I want to change that. Small changes in behavior add up in major ways. Helping each other out can actually save lives.