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.

What options are there for a policy study?

As a policy analyst, I encounter a range of different types of policy studies. Different research questions demand different tools. For example, measuring the economic footprint of a cycling industry requires an entirely different approach than figuring out whether data centers create net value for society.

Getting useful answers to policy questions starts with identifying what kind of policy study is necessary to answer these questions. Today, I wanted to talk about four different types of policy studies that we frequently encounter at Scioto Analysis.

Economic Impact Analysis

Economic impact analysis can be used to estimate how increased spending or employment ripples through the economy of a defined area. This type of analysis looks at how economic activity in one industry affects spending in other parts of the economy. While not policy analysis on its own, economic impact analysis can be a helpful tool for understanding the scale of an industry or project compared to the overall economy.

Economic impact analysis works by taking a baseline spending figure and estimating how that spending ripples throughout a regional economy. We do this using economic multipliers. For example, if someone purchases bread at a bakery, the business uses that revenue to buy more flour and pay its employees. Those employees then spend their wages on rent, groceries, transportation, and more. The flour miller uses the money from the baker to buy more wheat and pay its employees, and so on.

To actually do the analysis, we collect or estimate spending data and use tools like RIMS II or IMPLAN to calculate how much a dollar spent in one industry spurs new activity in others. These multipliers break into three layers:

1. Direct Effects: initial spending and employment

2. Indirect Effects: business-to-business interactions along the supply chain

3. Induced Effects: household spending from workers after paying into taxes and savings

However, economic impact analysis has some limitations. For one, an economic impact analysis does not establish creation of new economic value. A new sports stadium could spur hundreds of millions of dollars of economic activity on restaurants, retail, and ticket sales, but that consumer spending could have existed regardless of the sports stadium existing–just in a different part of the economy. Economic impact analysis also does not account for social costs, so it does little to tell us the tradeoffs associated with a certain policy or economic activity.

In January 2025, we released a study about the economic and health impacts of bicycling and trails in Iowa. This study is a great example of when to use economic impact analysis: we used survey data from Iowa cyclists to determine what economic contributions cycling had in Iowa’s economy.

You should choose an economic impact analysis when you need to demonstrate the size, reach, and regional economic contribution of an existing sector, institution, or new facility to local stakeholders and legislators.

Cost-Benefit Analysis

While economic impact analysis measures benefits and costs in the formal economy, cost-benefit analysis measures costs and benefits in the broader economy. Cost-benefit analysis is the systematic inventorying of all benefits and costs of a public policy, expressed in monetary terms. The goal of cost-benefit analysis is to determine whether a certain policy creates more total value for society than it costs.

Instead of looking at spending flows, a cost-benefit analysis assesses allocative efficiency as a whole. This means that cost-benefit analysis assesses nonmarket outcomes using empirical studies and comparing real-world conditions, instead of just looking at formal economic activity.

To conduct a cost-benefit analysis, analysts follow a sequence of steps:

1. Establish a baseline: Costs and benefits should be measured against a counterfactual scenario, typically the status quo.

2. Determine policy options: Costs and benefits should be compared across multiple discrete alternatives.

3. Decide standing: It should be explicitly clear whose costs and benefits count in the analysis, which typically involves geography.

4. Identify impacts: The analyst should have a list of impacts to analyze in cost-benefit analysis that is based on current research and literature.

5. Quantify and monetize impacts: All impacts should be assigned numbers and monetary values.

6. Discount costs and benefits: All impacts should be discounted to account for current costs and benefits being valued more than future ones.

7. Conduct sensitivity analysis: Key assumptions should be tested to understand the certainty of results and the importance of different assumptions.

8. Tell your story: Results are communicated clearly in a report, press release, presentation, or something else entirely.

Scioto Analysis publishes the State Handbook of Cost Benefit Analysis as a toolkit for state and local analysts looking to conduct cost-benefit analysis.

We applied this framework in our cost-benefit analysis of congestion fees in New York City. We monetized impacts such as driver time savings, emissions, fuel waste, consumer spending, and implementation costs to understand the net social impact of congestion fees in New York City. We found that in 2025, congestion fees in New York City generated approximately $2 billion in net social benefits.

You should choose a cost-benefit analysis when you need to decide whether a proposed policy or public project creates more value for society than it costs, or when you are comparing different policy options to see which one delivers the highest net social return to the public.

Cost-Effectiveness Analysis

Cost-effectiveness analysis answers a straightforward question: how much bang do you get for your buck?

While cost-benefit analysis focuses on dollars in and dollars out across society, cost-effectiveness analysis is a broader framework that answers a narrower question. It calculates total program cost divided by a specific, non-monetized outcome, telling you how much a policy costs to deliver a physical result without trying to assign dollar values to social outcomes.

By focusing on a single metric, cost-effectiveness analysis simplifies impact identification and gives agencies working under strict budget caps an intuitive way to compare policy alternatives head-to-head. Instead of quantifying many different impacts, cost-effectiveness analysis focuses on a specific metric. Cost-effectiveness analysis is a great way to compare different policy options that are trying to achieve the same goal. The main limitation of cost-effectiveness analysis is that it is, by nature, tunnel vision. While it does a good job of comparing different policies against a singular goal, it doesn’t account for unexpected impacts or opportunity costs.

You should choose a cost-effectiveness analysis when you have an established goal or a fixed budget and need to identify the lowest-cost way to reach that target across competing policy options.

Program Evaluation

Program evaluation is the retrospective, data-driven assessment of an existing or completed policy to determine whether it actually achieved what it set out to do.

While forward-looking policy research projects what might happen under a pending proposal, program evaluation deals with the hard data generated during or after implementation. Program evaluation requires us to ask: did the program itself cause the observed changes, and how large was that impact?

To answer that question, evaluators rely on an econometric toolkit designed to isolate program effects. For example, if participants in a job training initiative see their incomes rise, that alone does not prove the program succeeded. Exogenous factors such as regional wage growth or inflation could lead to rising incomes, while the job training initiative might have done nothing. Because of this, evaluators often use causal methods like difference-in-differences and regression discontinuity to create a counterfactual, an analysis of what would have happened if the program didn’t exist.

Because of the emphasis on causality, internal validity is key to program evaluation. To properly evaluate the effects of a certain program, we should approach our analysis with an objective eye that is free from selection bias or confounding variables.

You should choose a program evaluation when you have already implemented a pilot or program and want empirical proof of its causal impacts.

There are more policy studies beyond these, but thinking about policy analysis with these four frameworks is a good place to start.

How many Ohio data centers are subject to moratoriums?

Over the summer, I wrote a blog post discussing where Ohio’s data centers are headed. As of then, I calculated that there are currently about 143 planned or proposed data centers in Ohio. In some of these cases, there are more than 10 individual data center facilities planned for one campus. For example, in Shalersville, Bitdeer, a cryptocurrency mining and artificial intelligence infrastructure company, has 14 data center facilities planned in 2029 and 2031 for just one campus.

If all of these planned data centers are built by 2034, Ohio will more than triple its data center energy capacity. Increasing energy draw has implications on emissions, utility bills, and more. Many Ohio residents and municipalities share these concerns. According to a poll from Bowling Green State University, 78% of Ohio residents support data center moratoriums until more information can be gathered about their effects, 75% are opposed to construction in their local communities, and 63% are opposed to construction elsewhere in Ohio.

A great new story from the Ohio Capital Journal finds that at least 138 Ohio municipalities and townships have active moratoriums on data centers as of September 17, 23 communities have effectively banned data centers, with at least 23 more considering similar proposals. Today, I want to explore what would happen to Ohio’s planned data center projects under these moratoriums and bans.

How will moratoriums and bans affect Ohio data centers?

Of the 143 planned or proposed data centers in Ohio, 22 of them would be impacted by current moratoriums, one data center would be impacted by a proposed moratorium, and one data center would be impacted by a proposed ban. That leaves 119 data centers in Ohio that would currently be unaffected by moratoriums or bans across the state.

Part of the reason why relatively few proposed data centers in Ohio remain unaffected by moratoriums or bans is that the proposed data centers across Ohio are mostly concentrated in certain municipalities. For example, New Albany, Ohio has 42 proposed data center facilities across 15 campuses. And Jeffersonville, Ohio has 23 proposed data center facilities on the same Amazon Web Services campus.

The map below shows the planned and proposed data centers in Ohio and which facilities could be impacted by current and proposed moratoriums and bans.

Many of these moratoriums are set to end soon. The moratoriums in Niles and Shalersville are set to expire in November 2026, the moratoriums in Lordstown, Springfield, and Sunbury are set to expire in January 2027, and the moratorium in Urbana is set to expire in March 2027. Other moratoriums are expected to expire later though, with Commercial Point’s moratorium set to expire in November 2027 and Lima’s moratorium set to expire in March 2028.

The moratorium in Commercial Point is currently being challenged in court by an Amazon subsidiary. In May 2024, the Commercial Point Village Council moved to rezone large parcels of land to a district exclusively designed for data center development. The council originally voted 4-1 to bypass the standard public referendum period and pass the zoning ordinance as an emergency measure. By statute, however, the Commercial Point Village Council has a 6 member council, which was down to 5 members due to a recent resignation. 14 days prior to the vote, Commercial Point’s mayor indicated that waiving standard zoning ordinance rules requires 5 votes. Now, the Amazon Web Services data center’s existence comes down to whether passing the zoning ordinance for it was legal to begin with.

In the figure below, we show how annual carbon emissions would change if all current moratoriums and bans are maintained.

If current moratoriums and bans are maintained, we project that the annual electricity generation from data centers in Ohio will decrease by about 28 million megawatt-hours per year by 2028. For every 100 million megawatt-hours of electricity generation per year, we could power around 9.2 million homes. For reference, there are around 4.9 million homes in Ohio, meaning that by 2028, the electricity generation from Ohio data centers could power all of the homes in Ohio about 4 to 5 times.

What does the future hold for data centers in Ohio?

As opposition to data center development in Ohio grows, it’s possible that more moratoriums and bans will emerge, which could impact Ohio’s data center landscape even more. An interesting pattern I noticed is that many moratoriums and bans are happening in municipalities that border planned data center sites, but not within the municipalities data centers are planned for themselves. One possibility is that community opposition to data centers grows in areas that have planned data centers nearby, but it’s more challenging to pass moratoriums and bans in the municipalities where planned data centers are located due to fear of legal trouble or tax exemptions that have already been agreed upon.

Some data centers have begun implementing policies that could help ease community concerns. All of the data centers that are currently planned for Ohio have announced plans to implement closed-loop cooling methods. Closed-loop cooling is great for reducing water use, but it also uses more electricity than open-loop cooling.

Additionally, about 19% of proposed data center projects in Ohio are planned to use behind-the-meter electricity generation. This means that data center facilities would not be drawing electricity from Ohio’s statewide grid, which would lessen the impact of higher electricity prices on ratepayers. It also means that data centers might be using cleaner energy sources than the Ohio grid, which would reduce emissions.

As data centers grow both more popular among developers and more contentious around community members, it’s important to find policies that address community concerns around data centers. Scioto Analysis plans to release a cost-benefit analysis about different data center policies at the beginning of October where we analyze the impacts of different policies in Ohio such as a ban or requiring energy to be generated behind-the-meter.

What does adult literacy look like in Ohio?

Last week, I went to a webinar hosted by the National Coalition for Literacy and the Learning Impact Network about adult literacy rates. My big takeaway from the webinar is that adult literacy is a real problem in the United States. There are many highly educated people working to improve adult literacy rates, but the compensation in these roles is low.

Adult literacy is important for voting, employment, social mobility, and public health. Populations that struggle the most with adult literacy rates are incarcerated individuals, people living in poverty, individuals with learning disabilities, and non-native English speakers. For some of these groups, adult literacy could be the barrier between poverty and the middle class.

Adult literacy in the United States

58.9 million adults in the United States, or 28% of adults aged 16-65, perform at or below the lowest literacy level. This means that 28% of the working-age population in the United States have at best a basic sight vocabulary and can read short texts on familiar topics to locate a single piece of information. This is distinguished from the next highest level of literacy, which would indicate that individuals can read paragraph-length texts, compare information from text to what is already known, and draw straightforward inferences.

71.5 million adults in the United States, or 34% of adults aged 16-65, perform at or below the lowest numeracy level. This means that 34% of the working-age population in the United States can at best complete simple processes like counting, sorting, and basic arithmetic operations. This is different from the next level of numeracy, which would mean people can complete calculations with two or more steps or processes, such as percents and fractions or interpretation of simple data and statistics.

The future of adult literacy in the United States is not exactly promising. According to the 2025 results from the Programme for International Student Assessment, there has been a 14-point drop in reading performance among 15-year-olds in the United States since 2022. Reading performance in the United States is at one of the lowest levels since the Programme for International Student Assessment began in 2000. The United States has the largest reading gap between high- and low-performers among OECD countries, meaning that the disparity between the highest- and lowest-performing readers in the country is relatively high.

65% of adult educators hold a graduate degree or higher and 37% earn less than $20,000 per year for their work as adult educators. Most of the federally funded workforce for adult educators is part-time or volunteer, and over 80% of adult educators indicate that their program’s funding is not secure. Given that adult literacy rates might worsen if high school reading performance continues to decline, investment in adult education could be a low-cost way to improve human capital across the country.

Adult literacy in Ohio

So what does adult literacy look like in Ohio?

The National Center for Education Statistics hasn’t released updated state- and county-level estimates for adult literacy and numeracy yet, but they plan to later this year. For now, we can look at existing estimates for literacy and numeracy in Ohio from 2017.

As of 2017, about 18% of individuals aged 16-65 in Ohio perform at or below the lowest literacy level. This contrasts to 22% in the United States (or 28% in 2026). Adult literacy rates in Ohio are higher than adult literacy rates for the United States as a whole. Compared to other states, Ohio is middle of the pack. Ohio is the 23rd-highest state for level one adult literacy rates. The highest states are New Hampshire, Alaska, and Vermont, while the lowest states are New Mexico, California, and Texas.

The best county for literacy in Ohio is Delaware County by a significant margin, with just 6.5% of the working-age population at or below the lowest literacy level. Delaware County is home to multiple suburbs of Columbus. The county has relatively high rates of educational attainment compared to Ohio averages, and the median household income is $133,540, nearly double that of the state of Ohio’s.

The worst county for literacy in Ohio is Holmes County, with 37% of the working-age population at or below the lowest literacy level. Holmes County also has the highest concentration of Amish residents in the world. Compared to Ohio, Holmes County has low high school and bachelor’s degree attainment, but the median household income in Holmes County is $5,000 higher than the state as a whole. 

It’s possible that Holmes County’s high proportion of Amish residents means that literacy rates and educational attainment are relatively lower, while earnings do not trend downward in the same way. This could be because Amish residents make money and create social capital in ways that don’t require literacy, such as agricultural goods, crafts, and face-to-face interactions. Additionally, 49% of Holmes County residents speak a language other than English at home, compared to just 8% of Ohio. Amish residents speaking Pennsylvania Dutch rather than English could also be driving down literacy rates in the county. .

In terms of numeracy, Ohio performs slightly better than the national average. In 2017, about 29% of individuals aged 16-65 in Ohio perform at or below the lowest numeracy level. This is slightly lower than the  32% innumeracy rate in the United States as a whole (or 34% in 2026). Compared to other states, Ohio is once again middle of the pack, the 25th highest state for level one adult numeracy rates. The highest states for numeracy are New Hampshire, Minnesota, and Alaska, while the lowest states are Mississippi, Louisiana, and New Mexico.

Similar to literacy, the best county for numeracy in Ohio is Delaware County, with 12% of the working-age population at or below the lowest literacy rate, and the worst county for numeracy in Ohio is Holmes County, with 50% of the working-age population at or below the lowest numeracy rate.

The Ohio Department of Education and Workforce runs the Ohio Aspire program, which provides free services to adults preparing for high school equivalency certificates. They also offer services to gain skills that can increase employability and earning potential. In 2023, the Ohio Aspire program had a 23,290 student enrollment, including 13,996 in the adult basic education program (0-8th grade level), 545 in the adult secondary education program (9th-12th grade level), and 9,294 in the English for speakers of other languages program.

Of the program’s participants, 9,228 increased at least one academic level in 2023, 1,493 earned a high school equivalence diploma, and 2,554 enrolled in post-secondary education or training.

Adult education could be an overlooked tool for human capital development. People who can read and do math have more of an opportunity to get jobs and start businesses. The lack of investment in this means this could be an area of low-hanging fruit for policymakers looking to bolster local economies.

Policy analysis vs. advocacy: Why would I need analysis instead of more advocacy?

If your organization works on issues, you probably have people who can explain why they matter. You have communications professionals, lobbyists, coalition partners, subject matters experts, direct service providers: people who know these issues inside and out. If you know this issue is important, why bother analyzing it?

Sometimes, you don’t need to. If what you lack is a message, a means to communicate that message, or help mobilizing people to support your issue of choice, you probably need advocacy more than you need analysis.

Analysis will not tell you how to change hearts and minds. What it will tell you is what the evidence says.

What advocacy does well

Advocacy is a key part of a democratic society. When governors, legislatures, mayors, city councils, agency leaders, or city managers are making decisions about how to prioritize policy change, they are often making these decisions based on which advocates have got their ear. When policymakers are trying to understand a new policy they haven’t come across, they usually hear the first explanation from a lobbyist, not a peer-reviewed journal or a white paper.

Advocacy is also key for larger, community storytelling. Volunteers and organizers mobilize constituencies to contact policymakers and talk to the media about an issue. Good advocacy crafts compelling stories that are retold by journalists then repeated on social media and across dinner tables. These stories tie issues to values and persuade policymakers to act.

Policy changes because of effective advocacy.

Where advocacy is insufficient

This being said, advocacy alone can’t always lead to better policy. There are times when advocacy alone will be a weaker tool than advocacy working alongside analysis.

For instance, sometimes key audiences are skeptical. If you are coming in trying to convince a key policymaker, funder, journalist, agency, or stakeholder who already has a bias against an issue, independent analysis can be a tool for helping them understand the importance of the issue and how it impacts outcomes they care about. 

I think about my time in Nebraska when a bill to cover contraceptive services for Medicaid nearly passed a legislature that was hostile to contraceptive coverage due to a fiscal note from their legislative research office showing the long-term savings the state would make from investing in it. A skeptical policymaker audience was moved by a trusted, independent analysis.

Also, there are times when answers aren’t obvious. The child tax credit expansion in 2021 looked very similar to a plan put forth in the 2019 National Academies Roadmap to Reduce Child Poverty, a study undertaken because child poverty was a wicked problem that people did not have answers for. Luckily, that study was there when policymakers were ready to act.

Sometimes you need a number that can withstand scrutiny. The reason the health insurance numbers released during the debate over the American Health Care Act got so much attention was because they came from the Congressional Budget Office, an agency that had spent decades cultivating and defending its reputation as an independent analytical institution. Numbers calculated by an advocate on her own would not have carried the same weight.

And sometimes you just need to know if you’re right. Sometimes you need to learn about the intricacies of your policy and learn where you might be wrong. This seemed to be the deciding factor that led to the Reagan Administration backing the international Montreal Protocol ending ozone emissions worldwide: a cost-benefit analysis of its impact.

This is a risk around bringing an independent analyst in: you may not get the answer you expect to get. If you are looking for a study to reach a predetermined conclusion, independent analysis is probably not what you need.

What does analysis give me that advocacy doesn’t?

Analysis does some things very well that advocacy does not. Analysis lends credibility to an issue that advocacy on its own can’t command. Analysts can quantify issues, telling not just what direction an issue will push society, but how much it will, which is especially important when it comes to budgeting and regulation. Analysis helps identify and confront tradeoffs, which is a reality for policymakers but often ignored or overlooked by advocates. Analysis is also decision-oriented. While advocacy often focuses on whether policy is “good” or not, analysis focuses on when, how, and if a policy will work, along with what variations of a policy will work the best. Analysis also brings a defensible methodology to the table, forcing the conversation to happen on a different dimension than a conversation whose terms start with advocates.

An analysis done well will define the conversation in a different way than advocacy can alone.

When do you not need independent analysis?

Sometimes, you don’t need independent analysis. If your primary goal is to develop messaging or persuade supporters, if the relevant evidence already exists and you just need to communicate it, if you aren’t willing to acknowledge findings that might make you think differently about your issue, if your issue is primarily about abstract values rather than concrete outcomes, or if the cost of answering the question outweighs the importance of the decision, you might not need independent analysis.

How to decide

Analysis is not better than advocacy: they are two important and, in a functional democracy, indispensable tools for making good policy.

If you know what you want to say and you want to say it effectively, you need advocacy or communications support.

If you need to know whether a policy will have a desired effect, how large that effect will be, what the tradeoffs inherent in a policy are, or whether an argument you are making all the time will hold up to scrutiny, that is when policy analysis will be more useful.

If you are trying to figure out which of these situations you are in, I am always happy to talk about whether an analysis makes sense. Feel free to send me an email at rob@sciotoanalysis.com

What does an independent public policy analysis cost?

Public policy analysis is a professional service conducted by policy analysts at the federal, state, and local level. If someone is looking to engage with an independent public policy analysis firm, they can face a range of different levels of price and service depending on the depth of the analysis.

An independent public policy analysis can cost anywhere from several thousand dollars for a narrow analysis to over $100,000 for a large, data-heavy analysis project. For many analyses sought by nonprofits, associations, foundations, and local governments, having a budget from $20,000 to $75,000 is a realistic range.

This is a wide range. The reason it is so wide is because policy analyses can range from a relatively straightforward analysis of existing data to a statewide study involving primary survey and interview data collection, economic modeling, and scenario analysis.

What makes a policy analysis more or less expensive?

When scoping a policy analysis, there are a few analytical dimensions that can make the analysis more or less expensive.

An analysis with a narrow scope can be done more cheaply than an analysis with a wide scope. If a question is narrowly defined, analysts can spend less time combing over different areas of the literature and can find more direct ways to answer the question. On the other hand, a scope with many questions that interconnect with one another can lead to multiplying time needs and more elements to an analysis, which drives up costs.

Analyses with small geography can sometimes have lower costs than analyses with wide geography. This is because data collection can be easier on a small scale than on a larger scale. This can work the other way around sometimes, though. There are some datasets that are easier and cheaper to access at the national level than at the local level.

This suggests another cost driver for policy analysis: data availability. Publicly available data that is accessible is much cheaper to procure than data that sits behind a paywall or data that needs to be newly collected. When analysts need to do primary data collection like surveys, focus groups, and interviews, that can drive up costs even higher. It is nice to get new local data, but data collection can easily double or triple the budget of a project or even more. A well-designed randomized sample survey can easily drive an otherwise simple analysis into the six figure territory.

Even if a study is well-defined with a geography that matches easily-accessible data, sometimes methods can drive costs up. Simple descriptive statistics are much less time-intensive for an analyst to assemble than a best-practice cost-benefit analysis, economic impact analysis, forecasting, microsimulation, or other specialized modeling techniques.

Finally, presentation can drive up costs. If the project needs to be polished for public consumption, written to be read by many, presented to the public, designed with graphics, presented to stakeholders, or go through multiple revision iterations, that takes analyst time and resources and can drive up the costs of the analysis.

What ranges can I expect for the cost of an independent policy analysis?

These are some ranges you can think about for a policy analysis.

Focused policy analysis or technical memo: $10,000–$20,000

A policy analysis that is narrowly defined, using publicly available data to produce a concise report of a dozen pages or so, could fall within this range. Not all policy analysis firms will do an analysis of this size because it is often not worth the trouble for them, and a lot of analyses like this are done in house. Working with a social impact firm or a firm focused on small contracts is best for this level of analysis.

Standard independent policy analysis: $20,000–$50,000

This is a typical range for a policy analysis done by a firm like Scioto Analysis. Over the course of a few months, a group of analysts can take a well-defined public policy problem and use techniques like structured policy analysis, cost-benefit analysis, economic impact analysis, or microsimulation to answer a policy question. This usually requires phases of analysis to answer the question and technical rigor that could not be achieved on the minimum budget of a technical memo.

Major study with substantial original analysis: $50,000–$75,000

You’d usually get into this territory if you are trying to add substantial primary data collection to a standard independent policy analysis. Interviews and surveys can drive the cost of analysis up, and using tools like IMPLAN can drive up costs for equipment. Studies like this can also have extra costs associated with quality assurance since they are more complex.

Large, complex, statewide or multi-method study: $75,000–$150,000+

This is where you start getting into the territory that a university might be a competitive partner for a project. Since universities require large overhead costs for external projects, that can drive costs up. Projects like this can take multiple years and include extensive data collection, complex methods, and many levels of analysis and review.

How are policy analyses priced?

Firms price policy analyses in different ways. 

A common pricing approach is hourly. A firm will give an hour quote, then charge for the number of hours taken to do the project. This can sometimes lead to overages if a project ends up being longer than expected.

Another is pricing on a project basis. Project pricing means the client is paying for an analysis, not the analyst’s time. These are often paid half up front and half at the end, at project intervals, or at monthly intervals.

Another approach is payment by retainer. Having a policy analysis firm on retainer can help make quicker and simpler analyses more feasible because it eliminates the costs of getting the relationship started, which are steep for independent firms.

Hourly projects lead to the highest likelihood of unexpected cost increases, though they can also lead to savings in some cases. Retainer projects can sometimes lead to wasted hours, but keep transaction costs of bringing a firm on low.

Is a more expensive policy analysis better?

No. The relationship between value and price of a policy analysis is totally dependent on the question a public sector, nonprofit, or foundation client is trying to answer. Sometimes a low-cost policy memo is the best way to answer a given question. Sometimes a muti-year six-figure analysis is the best way to answer it. This depends completely on the question and how it interacts with the cost drivers listed above.

What should I budget for an independent policy analysis?

If you are planning to have a policy analysis done and do not have a detailed scope, budgeting between $20,000 and $75,000 will cover a large portion of policy analysis needed for governments, nonprofits, and foundations. Smaller studies can fall below that range while complex studies can exceed it substantially.

Before setting your budget, define the policy question you want answered, the geographic scope, the data needed to answer it, and how you expect to use it.

If you are trying to determine what kind of analysis fits your research question and budget, email us at rob@sciotoanalysis.com and we would be happy to talk you through a reasonable scope.

Majority of Ohio economists agree that local property tax abatements bring data centers to Ohio

In a survey released this morning by Scioto Analysis, 23 of 37 economists agreed that local property tax abatements are an effective tool for attracting new data center development in Ohio.

As elections approach, data centers have emerged at the forefront of policy debates. Some lawmakers are pushing for more data centers, while others are skeptical due to community and environmental concerns. In the past, Ohio has offered tax incentives to attract data centers, but other policy options–such as requiring behind-the-meter electricity generation, which would mean that data centers would have to build their own on-site electricity generation plans, or banning data center construction–have gained traction.

23 of 37 economists agreed that local property tax abatements are an effective tool for attracting new data center development in Ohio. According to Curtis Reynolds of Kent State University, “I think that [local property tax abatements] are effective for attracting the data centers. That does not mean that I think they are a good idea.” Among other economists who agreed, many shared the sentiment that tax abatements are an effective tool for attracting new data centers while also indicating that they are not a good idea. Of the remaining economists, 9 disagreed and 5 were uncertain. Roman Sheremeta of Case Western Reserve University disagreed, stating that, “Data centers locate where the power is, not where the abatement is.”

Economists were mixed on whether requiring new data centers to generate their energy behind-the-meter would prevent energy bill increases for Ohio residents. 16 of 37 economists agreed, 11 were uncertain, and 10 economists disagreed. Casey Boyd-Swan of Kent State University agreed, indicating, “In practice, energy costs may not increase as much for Ohio residents, but most on-site generation will be natural gas turbines, which concentrate both air pollution and around-the-clock noise in the community where data centers are located.” However, Lucas Engelhardt from Kent State University disagreed, explaining, “This just pushes the demand back a step. Instead of competing for energy itself - driving energy prices up, they’ll be competing for the resources used to produce the energy - driving up the cost of those resources.”

20 of 37 economists disagreed that requiring new data centers to generate their energy behind-the-meter will reduce the environmental costs of data centers in Ohio. According to Jonathan Andreas of Bluffton University, “Most of the new power generation is burning fossil fuels and it is impossible to burn fossil fuels without any environmental costs. Furthermore, this is a way for AI companies to evade environmental regulations because the EPA’s acid-raid regulations explicitly does not apply to the behind-the-meter power generation.” Of the remaining economists, 13 were uncertain and 4 agreed. John Conzelmann of Denison University was uncertain, explaining, “This comes down to substitution patterns, which seem to lean toward natural gas. But it also depends on Ohio’s current grid’s environmental costs, of which I am unsure.”

The Ohio Economic Experts Panel is a panel of over 40 Ohio Economists from over 30 Ohio higher educational institutions conducted by Scioto Analysis. The goal of the Ohio Economic Experts Panel is to promote better policy outcomes by providing policymakers, policy influencers, and the public with the informed opinions of Ohio’s leading economists. Individual responses to all surveys can be found here.

Should the state of Ohio ban property tax incentives for data centers?

Earlier this month, a group of Ohio House members largely representing urban areas of the state introduced a bill to prohibit property tax exemptions for data centers.

Property tax exemptions are a significant tool that can have a big impact on local government budgets. When evaluating whether prohibiting property tax exemptions for data centers are a good idea, it’s worth asking a few key questions that I like to ask about a range of policy issues: does it work, does it help, and who does it impact?

Okay, first, “does it work.”

The point of a property tax exemption is to give an incentive for a company to develop a project at a given site.

While older literature on economic development incentives says that only about 2-25% of economic development projects make their decisions based on incentives, newer studies that focus specifically on data center incentives imply this number could be closer to 33% for data centers.

While my inclination was to think that property tax incentives would follow the path of past incentive programs, recent analyses of data centers seem to be saying otherwise.

Now for “are they good.”

There are a few questions we need to answer for this. One is this: what is the opportunity cost of a property tax exemption?

Exempting a data center from property taxes means less money available for schools, county government, townships, and municipalities.

Even if data centers are highly sensitive to exemptions and one-third of them change decisions based on these incentives, this could still mean hundreds of millions of dollars in public funds left on the table for the two-thirds of data centers that aren’t making decisions based on incentives.

Another consideration is what the real social impact is of attracting data centers.

We know data center construction leads to more construction spending, wages, and capital investment.

So do subsidies for sports stadiums, but the economic consensus is that these subsidies are an economic development shell game rather than creation of new economic value.

Data centers could be similar: they could just displace other economic development that is bound to happen on its own.

Further social impacts that need investigation center on utility costs and environmental and public health.

Energy-heavy data centers will either drive up electricity costs and water costs and lead to burning of fossil fuels if they pull energy from the grid and water from public water systems or lead to construction of new gas-fired power plants if their energy generation is pushed behind the meter.

This will lead to higher carbon emissions and the degradation of air quality and associated asthma and lung disease that comes with emission of NOx, PM2.5, and SO2 into the air.

Lastly, the question of “who benefits” matters here.

Much of the benefits of data center construction will accrue to investors and data users who live outside of the state.

Carbon emission costs will generally be felt by people across the world as they exacerbate climate change.

On the other hand, both economic development benefits like jobs and public health costs like asthma and lung disease will be felt by local communities.

A final consideration is this: the state tends to give local communities wide range to use economic development incentives because different localities have different challenges when it comes to economic development.

A total ban on property tax exemptions for these purposes would potentially cut some communities off from economic development that could help them. 

But data centers are a big deal. It may be that making this policy at the state level is the most straightforward way to regulate an industry that is changing rapidly in the buckeye state.

This commentary first appeared in the Ohio Capital Journal.

Should we pay people who work in non-market settings?

Earlier this month, the Trump Administration moved to allow stay-at-home parents to collect childcare subsidies that had previously been earmarked for working parents. Essentially, instead of working parents receiving a subsidy to pay for external childcare so they can continue working, this change would allow parents to pay themselves to provide their own childcare.

In this blog, I don’t intend to talk about this policy proposal specifically. Instead, I want to look at the underlying economic question: how do we value work that happens outside of the labor market? 

How do we value non-market activity

The interesting concept at the core of this policy is the idea that people can and regularly do create social value outside of a traditional labor market setting. This can be challenging because the labor market is one of our best ways to compensate people for creating additional social value.

When people create social value outside of the labor market, it often goes uncompensated. Things like volunteering or caring for children at home increase social welfare even though nobody gets formally paid.

This also creates a measurement problem. If a family pays someone to provide childcare, that service counts toward Gross Domestic Product. If a parent provides that care themselves without payment, it generally does not. The Bureau of Economic Analysis produces separate estimates of household production to help account for this missing activity. 

Non-market activity is also one of the components of the Genuine Progress Indicator, an alternative economic measure to Gross Domestic Product. Including unpaid work helps us understand economic activity that market transactions alone do not capture.

Putting a dollar amount on unpaid work

One approach to monetizing nonmarket work is to ask what it would cost to hire someone to perform the same task. Economists call this the replacement cost approach. For childcare, we might use the compensation of a childcare worker to estimate the value of a parent’s time. BEA uses a similar approach, drawing on domestic workers’ wages to value unpaid household work. 

According to Independent Sector, the estimated national value of volunteer time was $36.14 per hour in 2025. That provides a useful benchmark for describing unpaid contributions, although it does not mean that every volunteer activity produces exactly $36.14 in benefits each hour. 

Another approach is opportunity cost: what someone gives up to spend time on unpaid work. Forgone earnings are one way to estimate this. However, replacement cost and opportunity cost answer different questions. One estimates the cost of hiring a substitute while the other estimates the value of the alternative use of someone’s time. Both are used in household production research. 

Imagine two parents providing similar care, one of whom could earn considerably more in the labor market. Their forgone earnings would differ, even if the care they provide is equally valuable. This is why the choice of valuation method matters. 

How this might work in practice

There is an interesting parallel here with how we regulate electricity distribution. Distribution networks act as natural monopolies where competition isn’t really feasible. Having several companies build competing sets of wires to the same homes can be more expensive than having one network. That limits the role competition can play in setting prices. Utility regulators work within this constraint. 

Under cost-of-service regulation, regulators assess costs and set allowed revenues, including an opportunity to earn a reasonable return on investment. In essence, the regulation attempts to replicate what we might expect the efficient outcomes to be in a competitive market. 

While this is not directly applicable to the case of trying to pay non-market labor (particularly stay-at-home parents), it does suggest that it may be possible for the public sector to step in and through regulation approximate what the outcomes of a competitive market might look like. 

One reason policymakers should consider pursuing something like this is if they believe that the social value provided by non-market labor is more valuable than the opportunity cost of lost labor market activity. If staying at home and providing high quality childcare is more valuable than one person working some other job, then it may be in the public interest to create a system that compensates that parent for choosing not to work in the traditional labor market.

Does the One Big Beautiful Bill leave SNAP policy up to chance?

Beginning in October 2027, the Supplemental Nutrition Assistance Program (otherwise known as SNAP and  formerly “food stamps”) is going to be significantly more expensive for states. Currently, the program is fully guaranteed by the federal government, so aside from some administrative costs this amounts to free money for the states. 

Now with the passage of the One Big Beautiful Bill Act, states will have to pay a portion of their SNAP benefits themselves. This change alone is potentially devastating to state governments. SNAP is a massive entitlement program which means that anyone who is eligible for benefits may claim them. This means that 1) some states might not be able to afford paying for any portion of their total SNAP benefits (especially in the face of balanced budget requirements) and 2) states wont know how much spending they are on the hook for until after the fact. 

Those two problems are significant enough that it might be enough to cause states to restrict eligibility or cancel the program altogether, but the actual implementation that states have to contend with stands to make matters even more complicated.

According to the new legislation, a state’s SNAP error rate will determine what percentage of the total costs they end up paying for each year.

What is the SNAP error rate?

The SNAP error rate is the percentage of total SNAP benefit payments that are estimated to have been incorrectly overpaid or underpaid to participating households. This is calculated by taking a sample of households that receive benefits, and auditing them to see if they received the correct amount of benefits each month.

There are two points about this calculation that are crucial to understand. First, this is about total error at the household level. Overpaying one household and underpaying another don’t cancel out in the aggregate estimate, they both count towards a higher error rate. The second point is that this is calculated on a random sample of households, and is subject to sampling variation. 

That second point is especially important because the new law uses this error rate to determine how much states will have to pay. Beginning in 2027, states with an error rate below 6 percent will continue to not have to contribute state funds toward SNAP benefits. States with error rates between 6 and 8 percent will pay 5 percent of benefits, states between 8 and 10 percent will pay 10 percent, and states with error rates of 10 percent or higher will generally pay 15 percent.*

How will this impact states?

A Brookings study published last month looked into state SNAP error rates and measured how much they vary from year to year. In short, they found that these cost-sharing thresholds are far too small to account for the natural variability of the error rate estimate. This is a clear example of why it is foolish to build policies around point estimates. Uncertainty that is inherent whenever we look at a sample makes it so states have basically no way to plan for how much their SNAP program will cost year to year.

Brookings estimates that the average standard error of a state’s annual SNAP error rate is about 1.1 percentage points. That means a typical 95 percent confidence interval spans roughly 4.4 percentage points. But the cost-sharing thresholds between 6, 8, and 10 percent are only two percentage points apart. 

To put that into perspective, imagine that a state has an estimated error rate of 8.5 percent. Under the new policy, that state would have to pay 10 percent of its SNAP benefits. But the statistical uncertainty surrounding that estimate means the state cannot reliably be distinguished from one with an error rate of 7.1 percent or 10.9 percent. Those three estimates would result in the state paying 5, 10, or 15 percent of its SNAP benefits, respectively, which could amount to hundreds of millions of dollars of differences in state required spending. 

When looking at the historical data, Brookings found that 43 states had error rates that moved between at least two of the cost-sharing categories between 2022 and 2025. Fourteen states moved between three categories, and North Carolina moved between all four. Only five states (Idaho, South Dakota, Vermont, Wisconsin, and Wyoming) had error rates consistently below 6 percent over that period. 

The potential dollar amounts involved are enormous. Using 2023 data, Brookings estimates that California, for example, could face anywhere from no state contribution to more than $2 billion in SNAP costs simply because of the statistical uncertainty surrounding its estimated error rate. 

This creates a strange situation where states can make substantial investments in improving their SNAP administration without knowing whether those improvements will actually reduce their financial responsibility. A state’s error rate can change substantially from one year to the next simply because a different set of households happened to be selected for review. 

The underlying issue is not that statistical sampling is inherently bad. Sampling is an essential tool for estimating characteristics of large populations, and every estimate based on a sample comes with some degree of uncertainty. The problem is using a noisy estimate to make a high-stakes financial decision without accounting for that uncertainty. 

States are required to balance their budgets, which makes unpredictable expenses particularly difficult to manage. A state cannot simply assume that its SNAP costs will be the same next year, because its required contribution could change by hundreds of millions of dollars based largely on statistical noise. 

If the goal of this policy is to encourage states to administer SNAP more efficiently, there are better ways to accomplish that goal. Making state governments responsible for billions of dollars in benefits based on an estimate that can change substantially because of random sampling variation does not create a particularly strong incentive for better administration. Instead, it creates uncertainty for state budgets and, potentially, for the people who rely on SNAP. 

*  There is a temporary exception for states with particularly high error rates during the first two years of the policy.