The challenge of categorizing people

Earlier this week, I attended the Minnesota Evaluation Association’s annual conference. In the morning, I had a conversation with a few other attendees about some of the challenges of performing quantitative analysis in a way that accurately reflects the groups we are trying to better understand.

At the heart of this conversation was how we as statisticians assign categories to certain questions and what the implications of those choices are. Grouping people together helps us identify patterns, but those same groupings can obscure important differences between the people we are studying.  

Categorical variables

At the heart of many statistical analyses is the question “how is X related to Y?” How do poverty rates differ across racial groups, for example? What happens, though, when we dig into the categories we use to answer that question?

Categorical variables allow us to organize observations into groups. Instead of looking at each person’s exact age, we might compare children, working-age adults, and older adults. We may choose to do this when looking at age depending on our specific research question, but for non-numeric variables we are often forced into studying them categorically.

One of the key problems with discrete categorization is that certain people might be grouped together on one measure like race or age, but that might not be a good representation of that individual. As an example, someone who is 64 might have more in common with someone who is 65 than with someone who is 25, even though our categories might place the first two people in separate groups. The boundary is useful for organizing an analysis, but it doesn’t necessarily represent a meaningful change in someone’s circumstances.

Tradeoffs with sample size

One way to address this problem is to use more specific categories. If a broad group contains people with very different experiences, we could divide it into smaller groups and analyze those separately. However, we face a tradeoff as we go down this path. The more specific our model, the less generalizable it becomes.*

Imagine we are using a survey to study employment barriers in a particular city. We might have enough responses to estimate an overall employment rate with reasonable confidence, but only a handful of responses from a particular immigrant community. The experiences of those respondents are important, but a few people’s answers may not give us a reliable estimate for the entire community.

In these situations, we sometimes have to combine categories or leave a separate estimate out of a report. This can be especially frustrating when the smaller population is the one we are most interested in understanding. A dataset can be large enough to answer questions about the broader population while still being insufficient for the central question of our project.

The important distinction is that we are leaving out an estimate because we cannot support it with enough confidence, not because the population is unimportant. We should make that limitation explicit. Otherwise, readers might interpret a group’s absence from a report as evidence that it faces no particular challenges.

Depending on the project, we might address this by combining multiple years of data, collecting additional responses from that community, or using interviews to better understand people’s experiences. Each approach has tradeoffs. Combining years makes it harder to describe current conditions, while interviews provide depth without necessarily telling us how common an experience is.

Tradeoffs with race data

At Scioto Analysis, we rely on the Census Bureau’s race categories for our analyses. Because we rely so heavily on the American Community Survey and other Census Bureau products, this is the only option we have. Even in situations where we collect other data, we often need to compare it to estimates from the American Community Survey, so we still need to use the same categories. 

The Census Bureau explains that its race data are based on self-identification and that its categories reflect social definitions of race rather than biological distinctions. These categories provide a common framework for analysis, but they cannot fully describe someone’s identity or lived experience.

Two people who identify with the same broad racial category might have different cultural backgrounds or speak different languages. Putting those people into the same bucket can hide those differences, which might be relevant to whatever policy decision we are studying.

At the same time, broad categories can still be relevant to understanding how people are treated. People with different backgrounds might encounter similar discrimination because others perceive them as belonging to the same racial group. For a policy question about unequal treatment, that shared experience may be important even when the category tells us relatively little about their individual identities. However, someone’s self-reported race does not directly tell us how other people perceive them.

The question, then, is whether the categories available to us capture the differences that matter for the policy we are studying. If we are evaluating access to a public program, language or immigration history might help explain barriers that a broad racial comparison alone would miss.

Quantitative analysis requires us to simplify complicated realities. Our models should increase our understanding of a topic, but we know that it will never be perfect. When dealing with categorical variables, especially with race data, we should be aware of the underlying assumptions that we are making by assigning categories, and we should be careful that those assumptions balance the tradeoff between understanding the data we have and preserving generalizability for the wider population.

* To see this tradeoff clearly, consider the trivial example where every individual is assigned their own category. Our analysis will give us plenty of information about each individual in our sample, but there is no way to make claims about people outside our sample, since their categories are by definition not included in the study sample.

What is “housing mobility” and why is it important?

Last week, I attended Housing Mobility 2026, a two-day conference hosted by Mobility Works in Cleveland. The first day was a training institute for housing mobility staff, especially workers from various local housing authorities around the country. The second day was a gathering of housing mobility practitioners from around Ohio and a presentation about what their work looks like.

I have two main takeaways from the conference. First, many of the people who work at local housing authorities care deeply about the families they serve, but many of them are overloaded. Housing authorities don't always have the time or resources to help each family as much as they would like. Second, many organizations outside of housing authorities are doing important work in the housing space. This work ranges from buying lower-cost market-rate housing to keep it affordable, to helping families find new homes in new neighborhoods and supporting them for several years after they move.

What is housing mobility?

Housing mobility programs help families with Housing Choice Vouchers, often called Section 8, move to neighborhoods with lower poverty rates and better schools. This assistance can include help with the housing search, moving costs, landlord outreach, and support after a move. Housing mobility programs often aim to improve outcomes for children, especially safety and education.

A voucher alone often isn’t enough for a family to move to a new neighborhood. Families face several other barriers to moving. Families may not know much about neighborhoods outside of their own, and searching for housing takes time that working parents might not have. Security deposits, application fees, and moving costs add up quickly. Moving to a new neighborhood might also mean leaving behind family, friends, and other support networks, especially if reliable transportation is a concern for families. Additionally, in most of Ohio, landlords may refuse to accept tenants with vouchers, though some Ohio cities have banned such “source of income” housing discrimination. Housing mobility programs try to address these barriers by helping families search for housing, covering upfront costs, and building relationships with landlords.

One study of housing mobility programs looked at the Moving to Opportunity program, an experiment the Department of Housing and Urban Development began in 1994. The program offered families living in high-poverty public housing vouchers and counseling to help them move to lower-poverty neighborhoods. Children whose families moved before they turned 13 were more likely to attend college, and they earned 31%, or $3,500, more per year in their mid-twenties than children whose families did not receive a voucher. As adults, they were also more likely to live in better neighborhoods and less likely to become single parents. Children who were older than 13 when their families moved did not see the same gains, which suggests that how long a child spends in a better-resourced neighborhood matters.

Last year, Scioto Analysis conducted a cost-benefit analysis of a Moving to Opportunity program in Ohio. We modeled an expansion of Families Flourish, a program in Columbus that gives families rental support, relocation help, and life coaching. We estimated a net present value of $149 million, or about $3 in benefits for every $1 in costs. Most of these benefits come from higher lifetime earnings for children, reduced crime, and fewer emergency room visits. Even with a higher 7% discount rate, the program produces $1.70 in benefits for every $1 in costs. 

What does housing mobility look like in Ohio?

About one in three Ohio households rent their home, slightly less than at the national level, and Ohio’s median rent is about $1,090 per month. According to the National Low Income Housing Coalition, Ohio has about 422,000 extremely low-income renter households, but there are only 37 affordable and available rental homes for every 100 extremely low-income renter households. 73% of these households spend more than 50% of their income each month on housing.

Organizations across Ohio are working on these problems from different angles. The Cuyahoga Metropolitan Housing Authority is one of the housing authorities in the United States Department of Housing and Urban Development’s Community Choice Demonstration, a national study testing whether mobility services help voucher families move to higher-opportunity areas. Early results from the Community Choice Demonstration across all sites found that 24% of families offered mobility services moved to an opportunity area within a year, compared to 4% of families who did not receive these services.

In central Ohio, Families Flourish, the program our cost-benefit analysis was based on, gives families three years of rental support along with life coaching to help them move to higher-opportunity neighborhoods. The program has served 130 families since it began, and reports that participants' incomes rose by an average of 58%, or $17,000.

Other organizations focus on housing supply. Enterprise Community Partners buys naturally occurring affordable housing, or older rental housing that is affordable without a subsidy, to keep it from becoming unaffordable. In Northeast Ohio, FutureHeights in Cleveland Heights and LakewoodAlive in Lakewood work to keep older homes in good condition. For example, FutureHeights focuses on finding older homes that would otherwise be demolished and improving them to act as affordable housing. LakewoodAlive helps many older homeowners who might otherwise not be able to afford improvements and repairs to their homes.

Housing mobility and housing supply programs approach the same problem from different directions. Mobility programs help individual families move to higher-opportunity neighborhoods. These programs are especially useful for voucher holders and for families who lack the time and resources to search for new housing independently. Compared to mobility programs, supply programs work at a larger scale by building, preserving, or repairing affordable homes, often in higher-opportunity neighborhoods. However, supply programs do not change which neighborhoods a family can access, and they don’t necessarily reserve affordable housing for low-income households. Often, housing supply programs are most useful for low- to moderate-income families who can afford rent prices close to the market rate.

Housing mobility could be an overlooked tool for economic mobility in Ohio. Children who move to better-resourced neighborhoods at a young age are more likely to go to college and earn more as adults. Investment in housing mobility could be an area ripe for policymakers looking to improve long-term outcomes for Ohio families.

Scioto Analysis releases report on human development in Ohio

This morning, Scioto Analysis published a report on human development in Ohio. We estimate Ohio’s human development index to be 88.3, which is 1.5 points lower than the national index of 89.8. Human development measures a community’s ability to provide key fundamental needs to its residents across three areas:

• Health. Ohio residents have an average life expectancy of 77.5 years.

• Standard of Living. Ohio residents have a mean individual income of $49,000.

• Education. An Ohio six-year old can expect to complete 14.6 years of schooling over her lifetime.

Human development is highest in the suburbs of Ohio’s largest cities. The area with the highest human development index is a group of suburbs near Cincinnati made up of Loveland, Montgomery, and Forestville, with a human development index of 96.5. Suburbs east of Cleveland and north of Columbus are also among the highest areas for human development in Ohio.

Human development is lowest in core urban neighborhoods and some rural areas. Urban Dayton has the lowest human development index in the state at 82.4. Central Toledo, East Cleveland, West Columbus, and Scioto, Adams, and Pike Counties are also among the lowest areas for human development in Ohio. 

Human development also differs across demographic groups. White Ohio residents are expected to live four years longer and earn about $18,000 more per year than Black Ohio residents. This is not mirrored in education, though. The gap between White and Black Ohio residents is less than one year, both in adult average years of schooling and in children’s expected years of schooling.

While Ohio has very high human development overall, human development estimates differ across regions and demographic groups. The full report can be accessed here.

How long does a policy analysis take?

One of the most important questions we have to answer when we are proposing a project to a client is “how long will it take?” This one question has major implications for project costs, it can inform other parts of a client's policy strategy, and it is the most important part of determining what a project’s scope will be.

Because this question is so important, I wanted to spend some time discussing it from the perspective of an analyst. In this blog post, I talk about why timelines are important, how we build our timelines at Scioto Analysis, and what happens when unforeseen circumstances cause timelines to change.

Why is timeliness so important in policy analysis?

Unlike academic research where the primary goal is to increase our understanding of a topic, the goal of policy analysis is to inform policy decisions. In some cases where a policy analysis is purely exploratory, this doesn’t make much of a difference. However, when there is an active policy proposal being considered, or if there is a date set for a vote on a policy, analysts can sometimes be under immense time pressure. The deadline is often determined by the policy process, regardless of how much time we might want to spend on the research.

This creates a tradeoff that all policy analysis must reckon with: longer projects may lead to more confidence in results, but shorter projects leave more time for results to be communicated and digested by decision makers. There is almost always another question we could investigate or another assumption we could test. The challenge is deciding when that additional research would meaningfully improve a decision and when it would simply delay getting useful information to the people who need it.

How Scioto Analysis handles timelines

While every project is unique, there are general principles that we have developed over our years of experience that help us when determining how long we should spend on a project. A useful timeline needs to account for the entire process, from gathering data to reviewing the results and explaining what they mean. Finishing the calculations is only part of finishing the analysis.

As an example, we tend to begin almost all of our cost-benefit analyses with a 12-week timeline in mind. Based on our experience, this gives us enough time to research a topic in sufficient detail, while still prioritizing getting results out quickly enough to allow time for the information to be communicated and absorbed. That starting point gives us something concrete to discuss with a client as we determine what their particular project requires.

When we are developing project timelines with clients, we try to strike a balance between time and depth. We know approximately how much time is needed to perform each part of a policy analysis sufficiently, while our clients know when they need a report completed and how detailed they want it to be. Identifying the most important questions early helps us build a scope that fits the time available. We try to balance these needs so that we are providing our clients with as much information as they need, within a timeline and budget that works for them. This not only creates the most useful reports for our clients, but also ensures that the policymakers who may make decisions based on this research have enough time to fully understand the analysis we provide.

What happens if there is an unforeseen complication?

While careful planning and preparation can reduce the risk of delays in a project, there always arise some moments where unforeseen circumstances crop up. Maybe the dataset you planned on using wasn’t available for all the years you needed it, maybe the whole project team caught the flu and can’t work for a week. When this happens, there are often two main options:

The first option is to extend the project timeline. This can work for those types of projects that are more exploratory or don’t have hard external deadlines that must be met. This option should only be considered if there is enough time for the audience of the report to receive and digest the information even if they receive it later. A policy analysis doesn’t do any good if policymakers only learn about it after they’ve made their policy decision!

The second option is to reduce the project scope. If there are extraneous parts of a project that the analysts and clients agree are not likely going to lead to a significantly different outcome, or if there are sections that are not directly related to the main policy question, then it could be more valuable to get the audience of a report less information in a more timely manner. The key is to preserve the work needed to answer the central question credibly. A narrower analysis can still be useful, provided we clearly explain what it covers and what remains unanswered.

Because Scioto Analysis is a small team, this risk is front of mind whenever we create a project timeline. As a result, we always like to include an extra flex week into our timelines in order to account for any delays that might crop up. For most projects, this ends up being an extra week of editing and polishing that makes the final reports even stronger, but it has come in handy for addressing project delays as well.

One of the advantages of being a small team is that we can be more flexible and adjusting to these challenges is easier than it might otherwise be. We have the ability to change around our internal workflows as needed to respond to these types of challenges. Because of this, we are extremely proud of the fact that in our nearly eight years of operation, we have not missed a major project deadline. Meeting those deadlines matters because our work is most valuable when it reaches decision makers while they still have time to act on it.

What does it really cost Ohio households to lose power for a day?

A couple months ago, I was spending time at a friend’s family cabin in Hocking County. When we awoke in the morning, the lights would not switch on. We got a call from his mother:

“Don’t open the refrigerator!”

When the power goes out, people lose more than electricity. We were lucky: we were on vacation with yesterday’s barbecue in the fridge. Often, people lose power and that means they can’t make it to work, they lose their week’s groceries, lose air conditioning during a heatwave, or have essential medicine go bad overnight.

When a customer pays her electricity bill, she is not only purchasing power. She is also purchasing reliable access to power. And she is making decisions based on that: what to buy, how to spend her time, trusting that when she plugs an appliance in and she flips a switch, she will have power.

Power companies don’t always have to live up to this, though. They often have to make decisions between grid reliability, service maintenance, and other priorities. And often, it is cheaper to leave customers with the costs of intermittent power outages.

Last month, a bipartisan group of 18 Ohio House members introduced a bill to change this.

Ohio House Bill 1014 would require utilities to provide electric bill credits to customers after sufficiently long outages at their homes.

After a 16-hour outage under normal conditions, a household would receive a credit of the greater of either $50 or their monthly bill plus additional credits for additional days, plus reimbursement for lost food and medicine. Businesses would be reimbursed with a less generous formula.

The bill also stipulates that utilities will not be able to recover those payments through higher rates to customers.

Essentially, the bill puts a price on grid reliability.

This payment is larger than some estimates of household willingness to pay to avoid outages.

According to analysis done by Lawrence Berkeley National Laboratory, residential customers have an average willingness to pay of $57 to avoid a 24-hour blackout. The average monthly electricity bill in Ohio was $135 in 2024 and is likely higher now.

But this isn’t just about compensating households in the case of a power outage. It’s also about preventing power outages in the first place.

If utility decision makers are mulling over whether to make grid improvements or invest in other changes, this could be a tool for encouraging prevention of power loss, which could lead to a more reliable grid overall.

Having a compensation rate at this level could also help improve equity of power outages.

The way the compensation scheme is designed, it makes the timeline for eligibility shorter for smaller blackouts. Restoring service to isolated or difficult-to-support customers can be expensive, while much of the cost of leaving them without power falls on those customers. Under this bill, these costs would show up in a utility’s bottom line.

Excessive fees could backfire, however. If the fee level is set too high, it could give incentives for utilities to spend more preventing outages than customers value, or divert resources from other grid investments.

Putting a price on outages, though, would give utilities a reason to account for costs that currently fall largely on customers.

This commentary first appeared in the Ohio Capital Journal.

Can we change the scope of a policy study along the way?

When we start a new project at Scioto Analysis, one of the first things we do with a client is agree on a scope: what questions the study will answer, which impacts will be measured, what methods will be used, and when the product will be delivered. But what if something comes up during the project timeline? Is it possible to change the scope along the way?

The short answer is yes. The scope of a policy study can change along the way, and sometimes it should. Changing the scope isn’t free, however, and how easy it is depends on the project. Today, I wanted to walk through what changing scope involves, when it makes sense to change scope, and how to plan for changes when they happen.

Changing scope has costs

When the scope of a policy study changes, some of the original plans around research questions, data collection, methods, and structure might need to change too. These changes usually show up in two places: budget and timeline. Adding a new impact to a cost-benefit analysis could mean reviewing new literature, finding new data sources, and quantifying new impacts. This requires analyst time, costs more money, and could push back the final deliverable date.

How disruptive a change is also depends on how long the project is. Short projects have little wiggle room, so there might not be space to absorb a new scope item without extending the deadline. At the same time, shorter projects are often less complicated, so adding to the scope could be relatively easy. Long projects have the opposite problem: over several months, a lot of the project is built on the original scope, and a late change could mean revisiting different parts of the project. However, longer projects might also have more time and flexibility.

There’s also a less obvious cost created by changing scope. Changes made late in a project leave less time to check and revise the new work. If a change of scope is rushed, new sections of a policy study could end up weaker than the rest of study, which is bad for decisionmakers.

Analysts should try to change scope as little as possible as a project progresses. It’s important to spend time at the start of a project clarifying research questions, deciding which impacts matter, and agreeing on deliverables. When a new question does come up, it’s worth asking whether it’s important enough to justify the added time and cost, or whether it would be better answered in a separate follow-up memo.

When is changing scope worth it?

Policy analysis doesn’t happen in a vacuum. Unexpected changes come up. For example, new data might become available halfway through a project. Or, a bill could be amended while you’re analyzing it. In these cases, sticking to the original scope could produce a study that’s less informative than it should be.

In our study of the economic and health impacts of bicycling and trails in Iowa that we published earlier this year, we originally focused on two things: the economic impact of cycling and its physical health benefits. Along the way, we decided to expand the scope to include mental health impacts and carbon emissions. This required us making changes to the scope and adding riders to the original project, but it ultimately generated insights that were helpful to advocates telling the story of the impact of cycling in Iowa.

When a scope change does happen, it's usually handled by amending the contract, increasing the budget, extending the deadline, or some combination of the three. Sometimes there's another option: trading one piece of work for another. If a client wants to add a new impact but can't increase the budget or extend the deadline, it may be possible to drop or scale back something else in the original scope. While this can curb some of the costs of the shift, it will seldom eliminate those costs because switching itself takes time, planning, and effort.

Plan for flexibility

The best way to make scope changes easier is to plan for them from the start. One way to do this is by building flexibility into the workplan. This could mean leaving some buffer time before the final deadline, adding a flex week or two, or identifying optional pieces of analysis up front that can be added if time and budget allow. That way, a change along the way doesn't have to interrupt the entire project.

Another approach, especially when changes are likely, is a more flexible retainer contract. Under a retainer, a client pays a firm on a monthly basis rather than for a single fixed deliverable. That keeps the firm continuously available to adjust scope, conduct additional analysis, or take on new questions as they come up. A retainer makes the most sense when a client expects their needs to shift, such as when they're tracking legislation that's still moving or managing a program that's still taking shape. 

For years, Scioto Analysis was held on retainer by the RISE Together Innovation Institute to support poverty analysis in Franklin County, Ohio. Scioto Analysis also had a long-running retainer with Ohio University’s Voinovich School of Government working on Social Return on Investment analysis. Both of these contracts were flexible enough that we could pivot as new needs and opportunities arose, making the friction of scope change much less costly.

At its heart, public policy analysis is an iterative process. It is different from policy research: analysts learn things along the way that prompt them to revisit earlier steps in the analysis process. So public policy analysis is more robust to changes in scope than more rigid research projects. The methodology is built to change along the way, though changes of course come with costs.

So, can you change the scope of a policy study along the way? Yes. But it’s important to plan ahead as much as possible, and changes to scope should only happen if necessary.

How housing supply impacts poverty

Housing is one of the largest expenses in most household budgets, which means that changes in housing costs can have a significant impact on whether or not a household is able to make ends meet. This is particularly important in places where housing supply has not kept up with demand.

Recent research from Zachary Parolin at the University of Oxford attempts to answer this question by looking at housing costs and poverty across the United States. The results suggest that in some of the country’s most expensive states, a substantial share of poverty can be attributed to housing costs. In Hawaii, California, Washington, D.C., New Jersey, Massachusetts, Colorado, Maryland, Connecticut, New York, and New Hampshire, between 16% and 34% of poverty is attributable to above-average housing costs.

How do we measure poverty?

Technically, housing costs should not impact poverty rates at all. This is because the Official Poverty Measure

only compares a household's pre-tax income to a poverty threshold that was originally based on the cost of a minimum food budget. The threshold varies based on household size and composition, but it does not adjust for differences in the cost of living across geographic areas.

This creates some obvious limitations. A household earning $30,000 in a relatively inexpensive rural area does not face the same financial constraints as a household earning $30,000 in San Francisco or New York City.

The Supplemental Poverty Measure attempts to address some of these by accounting for taxes and government benefits such as SNAP, as well as certain necessary expenses. It also importantly accounts for geographic differences in housing costs.

In practice, the Supplemental Poverty Measure tells us more about how a household’s income compares to their local cost of living, while the Official Poverty Measure tells us about how incomes vary across the country. 

Housing and poverty

Housing is included as part of the Supplemental Poverty Measure because it is an essential expense. Households often can’t simply stop paying for housing if the price goes up or if their financial situation changes. 

Parolin's analysis asks what would happen if rents were 20% lower while everything else remained the same. In the states included in the analysis, that hypothetical reduction in rents would reduce overall poverty rates by roughly 18% (DC and New Hampshire) to 26% (Hawaii) and child poverty by 21% (New Hampshire) to 38% (Colorado).

The study also mentions how the increase in housing costs over recent decades has also eaten into some of the gains from programs designed to reduce poverty. In the states included in the analysis, rising housing costs offset more than half of the increase in SNAP benefits between 1989 and 2023. In New Jersey, rising rents effectively consumed 84% of the increase in SNAP benefits.

How can housing policy impact supply?

With housing policy, there are a huge number of rules and regulations that make it difficult for state and local governments to simply build new supply. Things like zoning rules and building codes can make it difficult or expensive to add housing.

There is some evidence that changing these policies can have meaningful effects. Austin and Minneapolis, for example, have both seen inflation-adjusted rents fall by roughly 20% following housing reforms and increased construction.

These effects result from the fact that the housing market in a city is extremely interconnected. New housing does not just house the people who move into it, it also frees up the homes those people leave behind. Those homes can then be occupied by other households. As this process continues, adding new housing can create a chain of moves throughout the market, increasing the availability of housing at a range of price points. 

Of course, this doesn't mean that building a luxury apartment automatically makes housing affordable for a low-income household. The effects depend on how much housing is built, where it is built, and how responsive the market is to additional supply. In general though, housing supply affects the entire market, not just the people who move into newly constructed buildings.

What does this mean for policymakers?

A lot of anti-poverty policy focuses on increasing people's incomes. We give households cash or benefits through programs such as SNAP, the Earned Income Tax Credit, or the Child Tax Credit, or we increase wages through minimum wage laws. If we look at poverty through the lens of the Supplemental Poverty Measure, then we see that lowering costs is an equally effective way to reduce poverty. Housing is particularly important because it is difficult for households to adjust their consumption when prices increase. 

If policymakers make it easier to build housing in places where demand is high, the resulting increase in supply can reduce rents. Lower rents effectively increase the purchasing power of households..

If the price of housing is substantially higher because we have not built enough homes, then some households will be pushed into poverty simply because it costs too much to live where they live. One benefit of trying to tackle this problem via housing policy is that some options like zoning reform don’t require significant resources to enable.

That doesn't mean that every housing regulation should be eliminated or that building more homes will solve poverty on its own. It does suggest, however, that policymakers interested in reducing poverty should pay attention not only to how much money households have, but also to how much it costs them to live.

Scioto Analysis releases cost-benefit analysis of data center policies in Ohio

This morning, Scioto Analysis published a cost-benefit analysis of policies to regulate data centers in Ohio. We estimate that between 2013 and 2026, Ohio data centers produced a net social benefit of $62 billion, mostly from creating 96,000 new jobs and generating $75 billion in wages. However, as development accelerates, we estimate that data centers under current regulations will produce only $2.4 billion in net benefits between 2027 and 2034, with benefits only slightly exceeding costs.

If every planned data center project comes online by 2028, the number of data centers will increase by 50%, and their electricity use will more than triple. When data centers draw power from the statewide grid, electricity bills rise for Ohio ratepayers. We estimate that higher electricity prices will cost Ohio ratepayers $16 billion between 2027 and 2034. Because Ohio’s grid runs partly on natural gas and coal, data centers also produce substantial emissions. The largest costs come from air pollutants that harm local public health, which account for about $41 billion in costs between 2027 and 2034.

To show how policymakers can shape the future of data centers, we evaluated five policy alternatives between 2027 and 2034:

• Under the status quo, data centers would produce a net present value of $2.4 billion, with a benefit-cost ratio of 1.04. Counties see most of their job gains from their first data centers, so additional data centers add few new benefits while adding substantial costs.

• A ban on data center construction would produce a net present value of $35 billion, with a benefit-cost ratio of 2.7. Most employment benefits from existing data centers would remain, while emissions and utility costs would fall.

• A behind-the-meter requirement with 90% natural gas and 10% renewables would produce a net present value of $35 billion and a benefit-cost ratio of 2.2.

• A behind-the-meter requirement with 50% natural gas and 50% renewables would produce a net present value of $39 billion and a benefit-cost ratio of 2.5.

• A behind-the-meter requirement with 10% natural gas and 90% renewables would produce a net present value of $43 billion and a benefit-cost ratio of 2.9.

We conducted 10,000 simulations of data centers in Ohio with different variables to test our model. The status quo is the most likely alternative to have costs that exceed its benefits, and a behind-the-meter requirement with 90% renewables is the only alternative with a positive net present value in at least 95% of trials.

As data center development in Ohio continues to accelerate, policymakers have options to prevent costs from outpacing benefits. The full report can be accessed here.

Can Ohio do better than a gas tax suspension?

On Wednesday, the Ohio House of Representatives and the Ohio Senate approved a 90-day suspension of the state’s gas tax, planning to backfill it with funds from the state’s General Revenue fund. With a two-thirds majority in favor of the bill in each chamber, the bill will take effect immediately. The governor signed the measure into law Thursday.

The bill will cost the state an estimated $700 million according to the Legislative Service Commission. The bill itself sets aside $725 million in general revenue funds to cover these lost expenses.

While a majority of Ohio economists polled do not think a gas tax suspension will provide meaningful relief to Ohio households, this decision does reveal something: Ohio legislators are willing to pay $725 million to support households. What other ways could they do it?

Household rebate

One weakness of the current strategy is that a gas tax suspension only supports drivers. It is a mistake to say that the fuel price hike we have seen this year due to the War in Iran only impacts drivers. Fuel costs have driven up prices across the board and even households without cars are paying the price.

The current gas tax suspension disproportionately benefits two-car households and leaves Ohio’s 370,000 no-car households out, even as they face higher prices themselves. A change to a flat household rebate would mean each household would get about $150, with no-car households benefiting the same as 3+ car households.

Targeted rebate

This last point raises an important consideration: price increases are felt differently throughout the population. Low-income households are likely to feel the sting of higher prices harder than upper-income households. Targeting the rebate to the bottom quintile of households heaviest hit by increases would increase the per-household impact to $760.

Another option would be to target low-income households with children, the types of households that are likely to have the longest-lasting pain from economic shocks. Last year’s $450 million child tax credit proposal would provide up to $1,000 per household, providing more meaningful relief to more at-risk households at a much lower price than the gas tax proposal.

Targeted transportation relief

If lawmakers are particularly tied to transportation relief, there are ways to do that, too. County Job and Family Service offices already provide gas cards and bus passes to residents in need. Transit agencies like COTA have longstanding programs providing discounted transportation fares to people in need. Providing funds for county Job and Family Service offices and local transit agencies to temporarily expand transit options for low-income workers could concentrate relief in households most affected by rising transportation costs.

This time around, lawmakers opted for a simple, popular solution: temporarily suspend a gas tax to lead to a 90-day decrease in gas prices for every commuter, trucker, or nearby resident who fills up on gas in the state. Lawmakers showed they are willing to open their wallets in a moment like this. Next time Ohio residents experience a large, unexpected price shock they have the chance to be a little more creative and help people more at a lower price.

This commentary first appeared in the Ohio Capital Journal.

Consolidation vs. competition

Earlier this month, the Health Policy Institute of Ohio released a newsletter that focused on the percentage of outpatient healthcare facilities in Ohio that are independently owned. Currently only 31.6% of Ohio’s outpatient facilities are independently owned, the sixth lowest rate in the country. 

While proponents of this kind of consolidation suggest that it can improve efficiency by simplifying administrative barriers across different levels of healthcare, some research has shown that this has not borne out. Instead, mergers of this kind appear to increase healthcare costs overall while not meaningfully improving patient outcomes.  

If this kind of consolidation really is making healthcare more expensive without realizing any gains for patients, then it should be something policymakers care about. 

Why consolidation happens

From a theoretical perspective, we expect this kind of consolidation to increase efficiency when there are economies of scale. In other words, if the marginal cost of providing healthcare to an additional patient is sufficiently smaller than the fixed costs of maintaining an independent practice, then it could be better to have fewer providers serving more people. 

For example, a large healthcare system can spread administrative costs across more patients. There could also be benefits to having different parts of the healthcare system coordinate with each other, making processes like referrals smoother for patients. These kinds of circumstances create natural monopolies.

The most common example of a natural monopoly is for electricity distribution. It doesn’t make sense to have three companies each build their own set of power lines to every house in a neighborhood. The infrastructure is expensive, and duplicating it would not provide much additional value.

However, natural monopolies are still price setters rather than price takers, and unchecked they will still produce less than socially optimal quantities at a higher than optimal price, particularly for healthcare which we might expect to have an inelastic demand.

How can policy address this?

If healthcare does tend to act like a natural monopoly, then policymakers could look to the utility markets as an example for how to improve outcomes for people. Regulations in the utility market are designed to essentially simulate what the outcome would be for producers and consumers if the market was actually competitive.

Producers are guaranteed a return on their investment that is largely determined by what kinds of returns they could theoretically get in other competitive markets, and consumers get lower prices for electricity. The system isn’t perfect, but it largely does a good job of getting people electricity at reasonable prices, and it is almost certainly better than letting monopolies exert their full power. 

If policymakers did attempt to regulate healthcare similarly to how they regulate utilities, they would have to overcome a number of problems. First is the fact that although there is a lot of consolidation, healthcare is clearly not exhibiting the same kind of monopolistic tendencies that utilities have. There are still a large number of independent practices that are able to compete, but would likely be crowded out if the government got involved with larger regulated systems. 

Additionally, it would likely be prohibitively difficult to try to determine a fair price in all healthcare situations. A regulator would need to know what a healthcare provider's costs should be, what level of profit represents a reasonable return, and how much of the savings from consolidation should be passed on to patients or insurers. Healthcare is a much more varied service than utilities provide, and reaching a fair resolution for all sides would not be an easy task.

Still, this framework relies on the assumption that healthcare is a natural monopoly rather than a regular monopoly. In 2024, the Federal Trade Commission and Department of Justice began looking into healthcare mergers, specifically trying to determine if this type of consolidation was leading to higher prices without any noticeable benefit. 

The key difference between these two situations is that it is more efficient to allow a natural monopoly to exist with regulations, and it is more efficient to break up regular monopolies and allow competitive factors to drive efficient market outcomes.

Either way, policymakers should be aware that this consolidation in the healthcare industry is happening. It doesn’t appear so far as if this integration is increasing efficiency, but this could change over time. If it does, then it could be beneficial to try and create a framework that protects patients from high costs, otherwise it might require action to preserve the efficiency that comes from competitive markets.