When to Conduct Cost-Benefit Analysis: The Costs and Benefits of Cost-Benefit

In my work doing evidence-based policy analysis and extolling the virtues of cost-benefit analysis, I’ve had people ask me a question from time to time that seems quite esoteric and a little silly, but that any true cost-benefit analyst should care about: what is the return on investment for cost-benefit analysis? In other words, has there ever been a cost-benefit analysis done on cost-benefit analysis?

For the life of me, I have been unable to find a cost-benefit analysis done on cost-benefit studies. Cost-benefit minded policymakers, though, should be interested in finding an estimate of what the value of a cost-benefit analysis is. So let’s go through the exercise of what a cost-benefit analysis of cost-benefit analysis would look like.

First, let’s estimate the cost of a cost-benefit analysis. I am going to err on the more expensive side in order to more conservatively estimate the value of a cost-benefit analysis. According to Glassdoor.com, the average policy analyst in the United States earns a salary of about $68,000. Using a Boston Business Journal article from 2005 and adjusting to 2019 dollars, a high-end estimate of $27,000 for benefits and employment taxes (40% of salary) and an additional $2,300 for office space and equipment brings the total annual cost of a policy analyst to about $97,000. Let’s assume on the conservative side that the policy analyst will complete a cost-benefit analysis every quarter. This means that a cost-benefit analysis will cost, on average, about $24,000.

Now let’s estimate the worst-case scenario. This is a scenario where a policymaker commissions a cost-benefit study and decides to make the same decision she made before the study. Assuming there is no value to the knowledge that the policymaker is right, those $24,000 would go to “waste.” There would be $24,000 in accounting costs and zero dollars in benefits. From an economic perspective, we can use a high estimate for the marginal excess tax burden, again erring on the conservative side. Say the study was financed from income tax dollars and those tax dollars had an especially high drag on economic productivity, say reducing economic activity among those taxed by 43%. This means that the total social cost of a cost-benefit analysis would be about $10,000.

Let’s look at the flip side: an extremely successful cost-benefit analysis. This would mean that a cost-benefit analysis uncovered an especially inefficient program and identified an especially efficient alternative, which the policymaker then reallocated dollars towards. The most extreme example I could construct is pulled from programs listed in the Washington Institute for Public Policy (WSIPP)’s Benefit Cost Database. Let’s say there was a small “Fast Track prevention program” with 16 children participating. This program would have net costs of about $1.5 million to run according to WSIPP analysis (16 children x $91,000 net costs per participant). Say a policy analyst identified the loss and recommended the dollars be shifted to deploy an additional police officer using hot spot strategies ($430,000 in net benefits according to WSIPP) and this recommendation was adopted by the policymaker. This would result in a net benefit change in the policy of a total of $1.9 million.

This would mean the $10,000 economic cost of the study would be totally swamped out by the benefits of this program. As a matter of fact, with this ROI, the next 180 cost-benefit analyses could be “failures,” not showing new results or not leading to policy changes for other reasons, and this study alone would still lead to net benefits for the analysis program as a whole.

Let’s take a more moderate example. Let’s say a department of Juvenile Justice was reallocating funds towards higher-yield programs. What if a cost-benefit analysis found that an adolescent diversion system with services ($2,900 in net benefits per participant according to WSIPP) was having no better impacts than one without services ($9,500 in net benefits per participant according to WSIPP). This means a total of $6,700 of net benefits would be realized for every juvenile moved from the program with services to the program without, which means that if at least two juveniles were moved, the $10,000 cost-benefit analysis would have net benefits.

It’s worth noting here that the marginal cost of the implementation of cost-benefit insights is extremely low. While moving one participant from a services program to a standard program after a cost-benefit analysis would have net costs of $3,800, moving two participants would have net benefits of $2,900, moving three participants would have net benefits of $9,600, and so on.

With these exercises in mind, we can draw four rules for policymakers determining whether cost-benefit analysis is worthwhile in a given situation.

1. Use cost-benefit analysis when policymakers are open to changing their minds. If policymakers have a policy plan in place already and are unwilling to adjust it, a cost-benefit analysis is a waste of money and will not lead to policy change or net benefits.

2. Use cost-benefit analysis when current program effectiveness is suspect or alternatives may yield better returns. Opportunities to substitute low-benefit or negative-benefit programs for high-benefit programs can quickly justify the cost of a study.

3. Conduct cost-benefit analysis on large programs. Even the smallest changes identified by a cost-benefit analysis of a large program will justify the cost of the study. While small programs can justify the cost of a study with large enough changes, if total net program costs or benefits are expected to be only a few thousand dollars or less, cost-benefit analysis may not be a wise use of resources.

4. Conduct a variety of cost-benefit analyses to “diversify your analysis portfolio.” As the Fast Track example above shows, a killer cost-benefit analysis can justify scores of duds. If a policymaker is open to implementing more efficient policy, she should commission a number of analyses of specific, pertinent policy problems and empower staff to act on recommendations that come from these analyses.

Cost-benefit analysis isn’t going to pay off every single time it is carried out. But more cost-benefit analysis applied to a variety of large, previously unstudied programs will lead to better use of limited public resources with the right leaders at the helm. And this is why I’m spending my time on this sort of work, because when cost-benefit analysis is done right, it has the potential to be the most effective public policy intervention of them all.

Rob Moore is the Principal for Scioto Analysis.

Think Outside the Bus

I was at a community meeting last week where I heard someone suggest making the COTA bus system completely free.

While some at the meeting dismissed the idea as “prohibitively expensive,” it is worth noting that currently bus fares only make up one out of every five dollars of operating revenue for the COTA system. Making the system free would cost maybe $25 million, or 2 percent of total state and county funds. Not a negligible expense, but far from impossible.

Fiscal feasibility out of the way, the question then becomes what is the goal of free busing? The public has two main interests in subsidizing buses: transportation efficiency and transportation equity.

From an efficiency standpoint, cars exact costs on the public, as I wrote earlier this month, in the form of congestion, crash damage and injuries, emissions and infrastructure degradation. Subsidizing buses is an indirect way of reducing these costs; a more attractive public transit system will mean less people in cars, which will decrease congestion, crashes, emissions and road maintenance needs.

COTA’s shift from a coverage-focused transit plan to a ridership-focused transit plan represents a nod to this sort of thinking. By increasing frequency on key transit corridors, more people will bus rather than drive.

There are more efficient ways to reduce car usage, though. A per-mile tax on driving (whether through a vehicle miles traveled fee, as a recent Dispatch editorial suggested, or a gas tax, which approximates a per-mile charge) would encourage more walking, biking, busing and telecommuting — all options that reduce the negative impacts of driving.

In general, local governments have few options to assess direct fees on driving, so transit subsidies become a second-best option.

For a more efficient transit system, city and county leaders need to work with the state to find ways to enact vehicle miles traveled fees to bring the private cost of driving in line with the social cost. The recent Short North parking reform is one example of doing just this.

From an equity standpoint, busing subsidies are an indirect way to alleviate poverty. If most people on buses are low-income, a transit subsidy represents a transfer to those who need assistance most.

Unfortunately, this is an ineffective way to target those in poverty. A new bus line does little good for a family that needs child care, quality education, housing or food. The most effective way to help these families is to give them the power to make their own family decisions, and that means giving them cash.

This column originally appeared in Columbus Alive.

Ride-share tax could curb Lyft and Uber excesses

By Rob Moore

Last month, ride-share application Lyft released its annual economic impact report. In this report, Lyft concluded that riders spent an additional $17 million in Columbus due to the availability of its services.

Meanwhile, controversy swirled last month around studies suggesting that ride-sharing apps were decreasing use of public transportation across the country, and that Lyft and Uber were increasing congestion on college campuses.

Typically, the conversation around the impact of Lyft and Uber has split into two camps, mirroring many other public policy debates. In one camp are the people lauding the benefits of the new technology, and in the other are those who focus on its negative consequences.

As most issues go, the truth is somewhere in the middle.

Of course this new technology is creating value, as anyone who has spent 30 minutes waiting for a cab after the bars close on a cold Saturday can attest. Ride sharing also creates a lot of possibility for first mile/last mile solutions, access to medical care and reducing usage of public parking.

On the other hand, ride sharing creates a lot of the same problems that personal cars do. Another car on the road means more congestion. It also means increased emissions, an increased chance of injuries and fatalities due to car crashes and greater wear and tear on public roads.

The congestion problem has particularly reared its ugly head in New York City, where average Manhattan car speeds have slowed to six miles per hour during business hours.

Things in Columbus aren’t likely to get that bad. But more cars on the road still means longer commutes, more pollution, more crashes and faster infrastructure degradation.

Last year, NYC decided to cap the number of Uber and Lyft vehicles in the city to combat congestion. This is not a best practice of congestion alleviation: A survey of 40 prominent economists last year revealed that they overwhelmingly thought capping ride-sharing services would make residents worse off by arbitrarily restricting public use of ride sharing.

Instead, a more effective way to account for the costs of driving not borne by drivers and ride-hailers is via a congestion tax.

Last week, a new per-ride state fee for ride-share services in Manhattan went into effect. The policy is expected to reduce the number of cars on the road while raising revenue for improvements to public transit.

An even more efficient option would be for all cars to be taxed on a per-mile basis. This way, the negative impacts of personal cars are treated the same as the negative impacts of ride-sharing vehicles. Cars that are on the road more would pay more in proportion with the problems they create.

Unfortunately, local governments often have narrowly tailored authorization from the state to collect taxes, so a more limited ride-sharing tax might be easier to justify legally than a broader vehicle tax. On top of this, per-mile pricing presents some obvious logistical issues.

If Columbus wants to reap the benefits of ride-sharing while curbing its excesses, a tax is a better option than a cap.

This column originally appeared in Columbus Alive.

Franklin County could end extreme poverty tomorrow

By Rob Moore

Last week, Franklin County held the fourth of five meetings in its strategic planning initiative to address poverty in the county.

Up to this point, the planning process has been focused on gathering opinions of local leaders on four issues organizers of the meeting have suggested impact poverty: health, employment, housing and youth.

Following these meetings over the past few months, though, has got me thinking. What would it look like if the county tried to address poverty with a laser focus, working to find the most effective ways to reduce the number of people under certain income thresholds?

Let’s start at the very bottom of the income scale: extreme poverty. Believe it or not, there are still Ohioans who live on less than $2 a day, the global threshold for extreme poverty.

In 2016, the Center for Community Solutions in Cleveland, using American Community Survey data, estimated that about 190,000 Ohioans lived on less than $2 a day.

Assuming the rate of extreme poverty in Franklin County mirrors the ratio of deep poverty statewide, we can estimate that about 24,000 Franklin County residents are living on less than $2 a day, or 1.9 percent of the county population.

So how do we end extreme poverty? The simplest answer is to give people money.

Conversations about “basic income” have been raging in the policy community for a few years now. This is not a new idea: The U.S. House of Representatives passed a Nixon-backed basic income proposal as early as 1970. Currently, basic income pilots are being held in Oakland and Stockton, California, and city leaders have started a study committee for a pilot in Chicago.

What would a basic income program look like in Franklin County? Funding a program to push 24,000 Franklin County residents above the extreme poverty threshold would cost about $17 million at a 100 percent uptake rate. Assuming the uptake rate is 60 percent (on the high end for social services), the total cost of the program comes to about $10 million.

Franklin County’s main form of revenue is sales taxes. In 2016, Franklin County’s 1.25 percent sales tax rate raked in about $300 million. This means the sales tax would have to be raised by about 0.05 percent to comfortably fund such a program.

This means if you spent $100 shopping at Easton, you would be charged about 5 cents to fund a program that would end extreme poverty in Franklin County.

There are reasons to worry about this approach. Sales taxes are regressive, meaning they impact the budgets of low-income people more than high-income people. This means that people in poverty who are not in extreme poverty would bear a disproportionate portion of this tax, making it less equitable than an income tax, which the county cannot legally levy.

But it is exciting to see that a small sales tax increase would effectively end extreme poverty in Ohio’s largest county.

This column originally appeared in Columbus Alive.

Better Analysis Means Better Public Policy

In 1987, the Reagan administration signed onto one of the most important international environmental treaties of all time, the Montreal Protocol. This was an agreement for all of the countries of the world to phase out the use of substances that are responsible for ozone depletion.

Twenty-five years later, the Obama administration rejected a new proposed regulation to control ozone at the same time that it supported a regulation to control mercury.

For those readers who are savvy political analysts, you might be scratching your head. Why would the conservative Reagan administration bullishly take on new environmental regulation while the liberal Obama administration rejected it?

While a lot of factors came together to influence these decisions, there is one that had an especially large impact on both of them: administrative use of cost-benefit analysis.

Cost-benefit analysis is the gold standard for applied welfare economics. What does this mean? It means that the best way for us to forecast the economic impacts of a given policy is cost-benefit analysis.

The federal government has been using cost-benefit analysis for over a century now, starting with the Army Corps of Engineers using it to help design infrastructure projects during the Great Depression. A Democratically-controlled Congress first endorsed cost-benefit analysis in its 1969 National Environmental Policy Act and Ronald Reagan first instituted centralized executive cost-benefit analysis in 1969. This original executive order has persisted, with minor changes, through the H.W. Bush, Clinton, W. Bush, Obama, and Trump administrations.

Cost-benefit analysis is carried out by systematically identifying the economic costs and benefits of regulatory or tax and budget policy proposals then estimating these costs and benefits in dollar terms. Costs and benefits are then discounted to account for the opportunity costs of the dollars used and “sensitivity analysis” is performed to show how accurate the estimates are.

Policymakers find cost-benefit analysis to be valuable because it allows them to measure very different policies, ranging from tax to education to health, against each other quantitatively, giving them an idea of the magnitude of economic impacts these policies will have and who they will impact. Cost-benefit analysis is also helpful because the process unearths an important insight—what will actually happen when a policy is implemented.

Despite cost-benefit analysis’s popularity and long history at the federal level, states and local governments have been slower to adopt this valuable tool. A 2013 study by the Pew Charitable Trusts estimated that the average state only conducts two cost-benefit analyses per year. Ohio was on the higher side, conducting about four on average.

Use of cost-benefit analysis is on the rise, though. Leading the way are states like Washington, which has created a center at Evergreen State College, the Washington State Institute for Public Policy, to systematically analyze and catalogue cost-benefit studies for the state legislature.

Ohio has dipped its toes in cost-benefit territory legislatively and administratively. The Ohio Legislative Service Commission is a well-funded analytical agency supporting the Ohio General Assembly. The Legislative Service Commission does not conduct full cost-benefit analyses, but it does project direct and indirect accounting costs for the state for bills before the legislature.

The Ohio Common Sense Initiative is an embryonic form of the original Reagan Administration executive order centralizing executive cost-benefit analysis. The Common Sense Initiative does not conduct full cost-benefit analyses, but it does mandate “business impact analyses” aimed at estimated the impact of regulations on firms in the state.

As a social enterprise, Scioto Analysis works with policymakers directly to give them the analysis they need but is also committed to improving the quality of analysis in state and local government as a whole. In order to serve this second goal, Scioto Analysis will be releasing a series of studies and resources over the next several months around cost-benefit analysis, its prevalence, and its application, specifically in the state of Ohio.

Better analysis means better public policy. I, for one, am looking forward to seeing how much more effective, efficient, and equitable state and local government will make its policies as they continue to increase their use of cost-benefit analysis. Who knows: it might even be a place for a meeting of minds in our increasingly polarized political landscape.

Rob Moore is the principal for Scioto Analysis, a public policy analysis firm based in Columbus, Ohio.

Columbus housing supply not keeping pace with population growth

By Rob Moore

Columbus has enjoyed relatively cheap housing compared to other large metropolitan areas in the country.

American Community Survey data tells us that the average apartment in Columbus costs $916, which means housing is cheaper in Columbus than all but 13 of the top 50 metropolitan areas in the US.

But incomes are lower in Columbus than other major cities, aren’t they? So wouldn’t housing be more costly when factoring in incomes?

Actually, when factoring in income, Columbus fares even better compared with other metro areas: Only six of the top 50 metro areas have cost burden rates (percentage of renters paying more than 30 percent of their income on rent) below Columbus.

That being said, there are still a lot of Columbus residents who are spending more than 30 percent of their income on rent: 44 percent, according to American Community Survey data.

At the same time, the supply of housing is not keeping pace with growing population. Since 2005, Franklin County has gained more than 200,000 residents, a 21 percent increase over that period. At the same time, the number of housing units has only increased by 40,000, an 8 percent increase.

This means the population has grown two-and-a-half-times faster than the housing supply since 2005.

In order for the growth of housing in Franklin County to keep up with the growth in population, an additional 5,400 housing units would need to be built every year.

These are large numbers. The largest affordable housing developer in the country only started construction on 1,700 units nationwide last year. Developer mandates on affordable housing can make some difference on the margins, but they are not going to make up for the magnitude of this shortfall on their own.

Brookings Institution Analyst Jenny Schuetz has done fantastic work on housing markets that gives some guidance to local governments trying to ease housing supply worries.

Central cities gain much more housing from “reconfiguration,” such as modification of larger houses into apartments, or construction of accessory apartments in places like garages. Easing zoning restrictions that stand in the way of these reconfigurations could lead to more housing construction.

Zoning in general has been a culprit holding back the supply of housing units in many local markets. Many cities are still leaning on a mid-20th century vision of single-family houses when the market calls for more density, particularly near employment centers and transportation corridors.

The Minneapolis City Council made waves last month when it passed a comprehensive city planning ordinance that allowed three-family homes in residential neighborhoods, abolished parking minimums and allowed for high-density buildings along transit corridors.

While many urban attempts at housing reform have only tried to patch the problem (public housing) or force the rent price toothpaste back in the tube (rent control), Minneapolis’ approach takes the housing supply problem head on.

Minneapolis is a similarly rent-burdened city to Columbus (10th most affordable out of the top 50 metro areas). Columbus also has a fairly pro-development history. Maybe it’s time for Columbus to take a page out of Minneapolis’ book.

Rob Moore is the principal for Scioto Analysis, a Columbus-based policy analysis firm.

This column first appeared in Columbus Alive.

Don't Call It Economic Development

By Rob Moore

Earlier this month, City Council and the Franklin County Commissioners voted on separate proposals to commit $50 million each — a total of $100 million — to develop the property for a new proposed Downtown Crew stadium over the next 30 years.

This was a big win for the Crew. After a tumultuous year that almost ended in the charter Major League Soccer team moving to Austin, Texas, it appears that a deal has been struck to keep the team in Columbus conditional on a new Arena District development that includes a new stadium.

While saving the Crew is a big symbolic win for the city, some local officials have also praised the new stadium as an economic development tool. Unfortunately, this flies in the face of the evidence available about the impact of stadium subsidies on local economies.

A highly respected poll of economists conducted last year found 24 major economists agreeing that stadium subsidies will cost taxpayers more than the economic benefits they generate. Only one economist dissented. This isn’t quite “existence of climate change” level consensus, but it’s pretty close.

But don’t stadiums create jobs? Well yes, of course they do, but these jobs come from somewhere. Dollars consumers spend at a stadium are dollars they don’t spend at local movie theaters, bars and concert venues. All the empirical data available tells us that consumers don’t increase spending on entertainment when a new stadium is built, but rather shift their entertainment spending from other sources.

Now, economic development isn’t the only reason the city and county may want to spend public dollars on a stadium development. Maybe the $100 million is worth the symbolism of having an MLS sports team. Or maybe the $100 million is worth the 180 affordable housing units planned to be built on the development.

To put it in perspective, though, what else could the city and county spend that $100 million on over the next 30 years?

With $100 million, the city and county could give 100 unemployed people $15-an-hour jobs for 30 years.

With $100 million, the city and county could put over 600 low-income children in pre-k for free every year for 30 years.

With $100 million, the city and county could lift more than 1,000 families above the poverty line with cash transfers for 30 years.

With $100 million, the city and county could pull more than 2,500 minimum-wage workers paying market rate rent over the housing burden threshold (30 percent of their income spent on housing) with housing subsidies for 30 years.

With $100 million, the city and county could pull more than 4,700 Central Ohioans out of food insecurity every year by providing SNAP-ed nutrition programs for 30 years.

Some of these strategies would have economic development benefits. Others would achieve equity goals. The question we should be asking ourselves, though, is this: Does the symbolism of a new stadium Downtown outweigh the benefits of these other uses of public dollars?

This column originally appeared in Columbus Alive.

Rob Moore Discusses New Research on "Prognosis Ohio"

Scioto Analysis Principal Rob Moore recently appeared on Ohio health policy podcast “Prognosis Ohio” to discuss Scioto Analysis research on Ohio’s economy.

In the half-hour program, Dr. Dan Skinner, a health policy professor at Ohio University and host of the podcast, and Moore discussed the findings of Ohio’s Economy: 2009-2016, a report released by Scioto Analysis last month.

“The broadest…conclusion that comes out of this report is that income inequality is a drag on economic growth,” Moore said on the podcast, “that has direct implications for our public policy.”

Moore and Skinner also discussed the impact of state budget policy on public health.

“Money in people’s pockets and reduction of inequality and poverty also have health impacts,” Moore said, “if you can improve people’s health they will have better incomes, if you can improve people’s financial situations, they will have better health outcomes as well.

Prognosis Ohio is an Ohio health policy and politics report hosted by Dan Skinner.

Moore Appears on "Discussing Politics While Drinking"

Scioto Analysis Principal Rob Moore recently appeared on new Columbus podcast “Discussing Politics While Drinking.” The podcast is a political/comedic show hosted by Columbus entertainer Brylan May.

While on the show, Moore and May discussed the midterm election results, the state of evidence-based policymaking in Ohio, and political polarization.

“On average there are only about four cost-benefit analyses that happen in the state of Ohio every year,” Moore said on the show. “Meanwhile, we’re passing dozens of laws, hundreds of regulations at the state level, and we have 934 cities and villages that are passing tons and tons of ordinances.”

“Discussing Politics While Drinking” will have a rotating cast of guests discussing national and local political issues as well as a range of beers, wines, and spirit offerings on-set.

Weathering the next recession

By Rob Moore

The past decade has been pretty good for Central Ohio.

In 2018, the Columbus metropolitan area experienced its ninth consecutive year of economic growth. Over this time period, Columbus’ economy grew 3.1 percent per year on average, outpacing both the country and Ohio, which each grew at about 2 percent per year over that time period.

That being said, all good things eventually end. A recent poll by the National Association of Business Economics found that two-thirds of business economists in the U.S. expect a recession in the next two years.

Is Columbus ready for a recession?

This month, Columbus City Council is holding hearings on the 2019 proposed city operating budget. About two-thirds of the proposed budget goes to police and fire services, with the remainder going to parks, trash collection, health and human services, economic development and city administration.

When a recession hits, tax revenues go down and cities are forced to either raise tax rates or cut services.

Neither of these are good choices. Raising taxes further strains the wallets of workers who are already dealing with economic hardship. Cutting safety and human services can be unwise since homicide rates increase during recessions and temporary cuts to health and education programs have been shown to have long-term impacts on low-income families.

During the last recession, Columbus faced a shortfall of more than $100 million. In order to deal with this, the city passed a tax increase and avoided decreasing services.

In order to avoid this tradeoff during the next recession, the city has wisely grown its budget stabilization fund (popularly called the “rainy day fund”) over the past few years.

Currently, the budget stabilization fund stands at $76 million, or 8.6 percent of the general fund budget. It is supposed to climb to $81 million by 2020.

While this fund will fill some of the gap during the next recession, it may not make up the entire shortfall in city revenue.

Estimates for Ohio statewide revenue shortfalls say that a safe statewide budget stabilization fund should be somewhere in the mid-teens as a percentage of general revenues. At the city level, this would require tens of millions of dollars more than the fund has now.

On the one hand, you might say that Columbus’ fund can be smaller than the state as a whole because of the relatively stronger Central Ohio economy.

On the other hand, the city budget is 75 percent dependent on income tax revenues, which are more volatile than sales tax revenues. This means that city revenues may fall faster than the rest of the state, which has more of a sales tax base.

Given these factors, the city should consider increasing its investment in its rainy day fund. While it may be tempting to spend all available funds right now, going to the voters for a tax increase in the middle of the next recession, or cutting police services, health services and education services when they are needed most is a dilemma we don’t want to have to face in the future.

This column originally appeared in Columbus Alive.