What are “baby bonds?”

Last month, Scioto Analysis released its most recent calculation of the Ohio Poverty Measure. This measure calculates the total income of Ohioans, including benefits and after taxes and unavoidable expenses, and compares it to basic consumption needs.

This is how we usually measure poverty: through a lens of income. A problem that some economists have with this approach is that they believe that escape from intergenerational poverty also has to do with accumulation of wealth than income.

There is reason to believe this. While income disparities in the United States have been increasing, wealth disparities are even more pronounced.

A 2020 Pew Research Center analysis found that while the average high-income household makes about seven times as much as the average low-income household, the average high-wealth household has assets 75 times as large as a low-wealth household. This means the U.S. wealth gap is seven to eight times as large as the U.S. income gap.

Wealth does a few things for families. First, it provides a family safety net for working people. That “emergency fund” you keep in your bank account in case you experience a bout of unemployment? That’s wealth. Wealth helps working people weather the ups and downs of employment endemic to a market economy.

Second, wealth can be a source of income. Large amount of wealth invested can yield dividends that can provide income to a household. It also can be drawn on during retirement as an ongoing source of income.

Wealth also provides the safety that allows people to take risks. Starting a business or investing in a job that is unlikely to have large returns in the short-term is a lot easier to do if you have wealth to fall back on.

But how do we promote wealth, especially for families that are already living paycheck to paycheck and can’t afford to set money aside with pressing needs to pay for now?

Last month, Economist Darrick Hamilton published a commentary in TIME Magazine addressing this topic. His answer is a policy called “baby bonds.”

A baby bond works by paying someone a lump sum once they reach a certain age. This often comes from an investment made at birth that grows in value over time. These can be restricted in use or not.

A great example is Connecticut’s program, which automatically invests $3,200 for any child that qualifies for Medicaid at birth. Between age 18 and 30, the recipient can claim the fund and spend it on buying a home in Connecticut, starting or investing in a Connecticut business, paying for higher education or job training, or saving for retirement. The state of Connecticut estimates a typical bond will have a value of $11,000-$24,000 by the time it is claimed.

By targeting these funds to children who are Medicaid recipients, the Connecticut program focuses on low-income children. But a program like this could be targeted in a number of different ways: by being claimed by low-income households, targeted toward low-income zip codes, or even universal eligibility. Medicaid has the simplicity, though, of being a high-uptake program for low-income families that the state has good data on.

Income interventions are important for fighting poverty today, but wealth-based interventions like baby bonds could be a valuable tool for disrupting intergenerational poverty. State and local lawmakers interested in helping fight poverty in the long-term should consider policies like baby bonds if they are looking for creative ways to fight wealth inequality.

How does a policy analyst impute missing public benefits data?

Last week, Scioto Analysis released our updated Ohio Poverty Measure, a report that we’ve been working on since November. In this measure, we use publicly available data to understand the state of poverty in Ohio. Our methods are based on a wide range of other state and city poverty reports, all of which are heavily influenced by the Census Bureau’s Supplemental Poverty Measure.

To calculate the Ohio Poverty Measure, we primarily used data from the American Community Survey. The American Community Survey is one of the most useful datasets because it has a higher sample size than the Current Population Survey, which is used to calculate the Supplemental Poverty Measure. This makes it the best way to estimate what poverty looks like at smaller geographic resolutions. Though the American Community Survey has such a wide reach, it does have a few important drawbacks.

The most important limitation of the American Community Survey is that it doesn’t ask as many questions as other surveys do. It succeeds in providing detailed information about things like employment and income, but it doesn’t ask about things like medical expenses which we need to know for our poverty report. 

For this information, we turn to information in the Current Population Survey. The Current Population Survey is similar to the American Community Survey, but it asks a smaller number of people a larger number of questions. Here, the tradeoff is sample size for more detailed responses. 

While we could have used the Current Population Survey as the base data for our analysis like the Supplemental Poverty Measure, we’d be relying on a much smaller sample to make claims about all of Ohio. Since we performed our analysis at the Public Use Microdata Area level (the smallest identifiable geographic area in these datasets), this would subject our results to sampling error. 

So how do we use data from the Current Population Survey to fill in the missing data from the American Community Survey? Formally, this process is called “data imputation,” and there is a great deal of statistical research on the topic. 

There are many ways to conduct data imputation. One simple example is simply assigning every person the average value of a missing variable. In our context, this would be bad since something like medical expenses will be zero for many people and quite large for a small portion of people, though it does have the desirable characteristic that the imputed data will have the same mean as the original data. 

For the Ohio Poverty Measure, we follow the same steps for imputing missing data that other poverty reports before us have. We use a two-step modeling process to first determine who is likely to have non-zero missing values, and then isolating that group we try to determine what the value is. 

To do this, we build two regression models from the Current Population Survey data. The first is a binary outcome regression that predicts the probability of an individual response having a non-zero value. The second looks only at those responses that have non-zero values and predicts the size of the missing variable. 

We then take these two regression models and use the American Community Survey data to get predicted values for the probability of a non-zero value. We then estimate the total size of the missing variable. 

Then, we rank the American Community Respondents by their predicted probability of having a non-zero missing value. We want to make sure that the same percentage of people in the American Community Survey have non-zero values as in the Current Population Survey, so we only count the most likely people until the proportion in the American Community Survey matches that in the Current Population Survey.

Making predictions is one of the most important parts of policy analysis. We often think that the predictions are the outputs of our work, not part of the input. However, with some clever statistical thinking, we can give ourselves access to really amazing data like the American Community Survey, even if it doesn’t have exactly all the information we need. As long as we can find a good way to impute it, we can take advantage of everything else it has to offer.

Is it “cost-benefit analysis” or “benefit-cost analysis?”

At Scioto Analysis, we are doing a multi-year project where we are demonstrating how a good cost-benefit analysis is conducted. 

As part of this series, we have conducted cost-benefit analyses on the state Earned Income Tax Credit, school closings for COVID-19, AmeriCorps, urban canopy programs, water quality programs, an Ohio child tax credit, legalization of medical marijuana, and daylight saving time. We are currently conducting a cost-benefit analysis of a minimum wage increase for Ohio.

We also are members of the Society for Benefit-Cost Analysis, the international association of analysts across the public, private, and academic sectors working to improve the theory and practice of benefit-cost analysis and support evidence-based policy decisions.

That’s not a typo: we conduct “cost-benefit analysis” and are members of the Society for “Benefit-Cost Analysis.”
So what is the difference between “cost-benefit analysis” and “benefit-cost analysis?”

Nothing.

These two phrases are used interchangeably in the world of cost-benefit analysis and are often used by different people, but refer to the same phenomenon.

The main differences between the two phrases are where they are used. “Benefit-cost analysis” is common in academia and in federal regulatory decision making. “Cost-benefit analysis” is more common outside of these sectors in the United States and in non-U.S. contexts.

But why do two phrases refer to the same practice? Below are some of the explanations I have heard over the years. I will be clear: I can’t vouch for any of these. I don’t know how true any one of these are, but they have nonetheless been offered to me as explanations for why people say “benefit-cost analysis.”

To represent “professionalized” practice

One explanation relayed to me by my colleague Michael Hartnett from the most recent Society for Benefit-Cost Analysis conference was that cost-benefit analysis had a push for mainstream acceptance in the 1970s, before Ronald Reagan required cost-benefit analysis of all federal regulations. Economists were trying to standardize the practice and promote it as a systematic form of applied economic analysis. In order to differentiate the practice from more sophisticated approaches to evaluating policy, “benefit-cost analysis” was put forth as a way to refer to the systematic practice.

To emphasize the importance of benefits

“Cost-benefit analysis” seems to focus on costs before benefits because…it comes first in the phrase. By placing the word “benefit” first, “benefit-cost analysis” assuages the fear of people who think conducting this analysis is overly focused on costs policy to the detriment of its benefits.

This explanation sounds a little silly, but it does fit with some worries people have. The loudest voices against cost-benefit analysis are often advocates who are afraid costs of policies they champion will outweigh their benefits. This theory is that by placing benefits first, those people will have their fears assuaged.

To reflect the formula of “benefits minus cost”

The central formula of cost-benefit analysis is calculation of net present value, or 

Present Benefits - Present Costs = Net Present Value

By placing “benefit” first in the phrase, we capture that central formula in the technique. This explanation is similar to the previous one: it is about trying to get people to understand how the system works. Seems a little weak for the confusion created, though.

Linguistic cadence

This is an especially interesting one: that the phrase “cost-benefit analysis” rolls off the tongue better than “benefit-cost analysis,” so “cost-benefit analysis” will persist no matter how much people try to get others to use the latter. The argument has to do with word emphasis within the phrase. I don’t know how true this is, but it is interesting.
Overall, the battle between “benefit-cost analysis” and “cost-benefit analysis” seems a lot like the battle between the phrases “this data” and “these data,” classic linguistic squabbling, sometimes between elites and mundane use, rarely important. While I will not be soon to give up my membership at the Society for Benefit-Cost Analysis, we’ll probably continue using “cost-benefit analysis.” Why? Because that’s what policymakers tend to use, and we’d rather have them understanding the analysis than reading articles like this.

Ohio Poverty Measure finds over 260,000 Ohioans pulled out of poverty by Social Security

Today, Scioto Analysis released its updated Ohio Poverty Measure, an indicator specifically tailored to estimate the extent of poverty in the state of Ohio. Using this measure, we find that in 2021, 8.7% of Ohioans lived in poverty. This is lower than the 12.1% poverty rate according to the Official Poverty Measure and higher than the 8.1% poverty rate according to the Supplemental Poverty Measure, the two main poverty measures calculated by the United States Census Bureau.

Among public benefit programs, we estimate social security has the largest impact of any public benefit program in Ohio, lifting over 260,000 Ohioans out of poverty in 2021. The measure also finds SNAP benefits, formerly known as “food stamps,” had a substantial impact on poverty, reducing the statewide poverty rate by nearly two percentage points.

The Ohio Poverty Measure is the most accurate measure of poverty in the state, using methodology inspired by the California Poverty Measure, New York City Poverty Measure, Oregon Poverty Measure, and Wisconsin Poverty Measure. The Ohio Poverty Measure was first calculated by Scioto Analysis in 2021, using data from 2018. This report constitutes the first comprehensive update of that data, giving estimates of poverty from 2021.

The Ohio Poverty Measure estimates the impacts of government assistance, the tax system, and expenses based on geographic cost-of-living differences. Including these adjustments makes the Ohio Poverty Measure more precise than both the Official Poverty Measure and the Supplemental Poverty Measure.

According to the Ohio Poverty Measure, Black Ohioans are 75% more likely than White Ohioans to be experiencing poverty, with 14% of Black Ohioans experiencing poverty compared to only 8% of White Ohioans. 

Additionally, we also find stark geographic disparities in poverty rates. Ohio residents living in urban core geographic areas and rural Appalachian communities experience poverty at much higher rates than those across the state as a whole. Ohio residents living in suburbs surrounding Ohio’s largest cities experience poverty at much lower rates than residents across the state as a whole. 

Child care in Ohio must promote both equity and quality

At Ohio Gov. Mike DeWine’s State of the State address earlier this month, he announced a voucher program to help low-income Ohio families pay for child care.

DeWine’s current proposal is to provide vouchers for families making up to 200% of the federal poverty level, $60,000 or less for a family of four. This follows in the footsteps of other proposals to target child care assistance to low-income families.

I’ve written before about how child care is a double-edged policy problem. Often when we talk about child care, especially from an economic development perspective, we think about it as a way to free up time for parents and allow them to work. But child care arrangements are also the place where children spend substantial time during some of their most important years of development. 

Because of the amount of time spent in child care, the quality of child care arrangements can have significant impacts down the road for child development.

A key problem with the child care system is that parents often need to prioritize child care that fits their work needs. Young children get no say in the child care arrangements they take part in and don’t have the capacity to evaluate which child care arrangements will lead to better outcomes for them down the road.

This is a central problem with child care Ohio State Professor Emeritus David Blau writes about in his book, “The Child Care Problem: an Economic Analysis.” Since parents are making decisions for children, they often underconsume high-quality child care.

This is why Blau recommends a system in this book that does two things: provides subsidies to low-income families to handle the equity problem of child care while giving higher subsidies to high-quality centers to handle the above efficiency problem.

Quebec had an experiment that shed some light on what happens when only one of these two levers is pulled. In the late 90s, the Canadian province enacted a universal child care program that charged families just $5 a day for child care. Policymakers declined, however, to tie subsidies to child care quality.

An evaluation of this program after 20 years conducted by well-regarded economists found the program led to a sizable shock in outcomes for children. Not only did this program lead to negative effects on non-cognitive outcomes that persisted into school years, it also led to worse health, lower life satisfaction, and higher crime rates for these children later in life.

Ohio has been bandying about different proposals around child care quality for years now. The current law of the land is the Step Up to Quality program, a five-star child care program that provides higher subsidies to higher-quality programs. Last month, the Ohio Department of Children and Youth and Ohio Department of Job & Family Services proposed a new rule to reduce this five-tier rating system to a three-tier rating system which could reduce payments for quality.

Whatever the state settles on when it comes to payments for low-income families and centers, Blau’s recommendations still ring true: subsidies should be targeted toward low-income households to advance equity concerns and should be targeted toward high-quality centers to support child development. This ensures both low-income families today and children tomorrow are benefiting from public investment in the child care system.

This commentary first appeared in the Ohio Capital Journal.

How do we incorporate animals into cost-benefit analysis?

A book I’m particularly excited to read that came out earlier this year is Dog Economics: Perspectives on our Canine Relationships. This book, written by two of the leading economists in the public policy field, applies a range of economic concepts to dogs and their relationships with human beings.

I first was exposed to David Weimer’s work on dog economics when I was serving as editor for On Balance, the blog for the Society for Benefit-Cost Analysis. Weimer had recently won the Society’s Journal Article of the year award for an innovative analysis he had done to estimate the value of a statistical life for dogs.

Why calculate the value of a statistical life for dogs? Well the reason is because federal agencies sometimes promulgate regulations that will directly impact canine well-being. Weimer was drawn to this idea because the Food and Drug Administration had conducted a benefit-cost analysis on approval of a new type of dog food. A glaring omission from the analysis? The risk to health for dogs who consumed the product.

But how do we calculate the value of safer dog food? When human beings participate in the labor market, they trade off risk of death for more pay. Using these tradeoffs, we estimate how much people value marginal reductions in risk of death, which allows us to come up with the value of a statistical life.

Dogs don’t participate in the labor market. In a literal sense, dogs don’t participate in markets. They don’t have income, they don’t have assets, they don’t purchase goods and sell their services. So how can we estimate a dollar value they put on reductions of their risk of death?

Weimer’s answer is to ask their owners. Weimer conducted a “contingent valuation” study to see how much value people put on their dogs’ lives. He did this by posing questions about hypothetical vaccinations for dogs that would reduce dogs’ likelihood to contract potentially deadly diseases. By eliciting the willingness to pay dog owners have for these vaccinations, Weimer was able to derive a value of statistical life for dogs.

This is certainly a step in the right direction for incorporating the value of public policy to animals into cost-benefit analysis. But this approach still falls short of fully incorporating animals into cost-benefit analysis. Why? Because it defines the benefits that accrue to animals through their owners’ altruism rather than on their own terms.

This would be sort of like if you surveyed people on how much less income they’d be willing to take on as a family in order for their spouses to have the opportunity to take on less dangerous jobs. It should be a good proxy for the value of a statistical life, but a more accurate measure would come from seeing how much people actually take on themselves, not relying on a report from someone who cares about them.

This approach is taken often when assessing benefits to children. But even here we see problems. In his book The Child Care Problem: An Economic Analysis, economist David Blau talks about the market failure caused by parents underconsuming high-quality child care. This happens because parental demand for child care that will nurture their children and lead to better outcomes for them and society falls short of the optimal social benefit. If children could rationally decide for themselves what quality child care to consume, they would likely be willing to pay more for quality child care than parents do.

For now, though, it is difficult for us to divine how much of a value of risk of death dogs take on themselves. Some cutting-edge researchers are conducting studies with chickens, seeing how much feed they are willing to give up in order to live in more free-range environments. The results are promising: initial studies have found chickens exhibit a rational demand curve for more open space.

Non-human animals are not as different from humans as we act like they are. They have to operate under conditions of scarcity just as much as humans do. And they are often subjected to the impacts of public policy just like human beings are. Hopefully as we innovate new ways to study public policy, we will also find ways to incorporate their interest into our benefit-cost models.

How do immigrants impact the labor market?

It’s presidential election season again, which means that debates around the most politically divisive issues are front and center in the national news. Immigration is one of the most commonly debated topics, and according to data from Pew Research, 57% of Americans think that it is a top priority for the President and Congress. 

Political narratives aside, what actually happens when immigrants come to the United States? You will get extremely different answers to this question depending on who you ask, so let’s instead use this opportunity to explore some of the data surrounding immigration. Specifically, what happens to the labor market when immigrants move in. 

Before we talk about actual data, it will be useful to visit some of the basic economic theory behind this question. If a wave of immigrants move into a community and begin participating in the labor market, then theory says we should see a growth in the labor supply. Assuming that nothing else changes, this would have two main effects.

First, unemployment should rise. Assuming that the demand for labor is unchanged, then there would just be more people competing for the same number of jobs. 

Second, wages would go down. Competition among potential workers should cause a race to the bottom for wages. 

These two outcomes only happen in a perfectly competitive labor market, where all workers are competing for the same jobs and wages can fluctuate with supply and demand. In the real world, the labor market functions very differently. 

To see just how differently this plays out in real life, we can look at the example of the Mariel boatlift, a mass migration event where thousands of Cubans moved to Florida in 1980. This event was the subject of a landmark 1990 paper by the economist David Card, who explored the labor market impacts of such a sudden shock.

Card’s study was a pioneering deployment of “difference-in-differences” economic evaluation. Card looked at unemployment and wages in Miami before and after the Mariel Boatlift then compared these to changes in unemployment and wages in comparable metropolitan areas. These comparable metropolitan areas served as “synthetic controls,” mimicking an experimental design and allowing Card to see what the actual impact of the sudden influx in workers had on employment and wages

Card found that, when compared to comparable metropolitan areas, the sudden wave of immigration actually had no effect on unemployment or wages in low-skill industries. Instead, the Miami labor market was able to immediately grow and absorb these new workers. 

Why was this able to happen? The dynamics of labor markets are extremely difficult to understand, but I think there are a few main reasons.

One reason is that in a healthy economy, there should be room for growth. Many economists consider there to be “full employment” not when the unemployment rate is 0%, but rather when everyone who wants to work is able to with relative ease (the rule of thumb I learned is that full employment is equivalent to an unemployment rate of about 3% - 4%). There are far more technical definitions of full employment, but the idea is we should always expect there to be some job openings. 

The second reason is that immigrants don’t just come here to work, they come here to live. They spend money in their new communities, they start their own businesses, they grow the local economy. By moving to Miami, the immigrants from the Mairel boatlift created the opportunity for employers to expand their operations. 

The political discourse around immigration isn’t very data-driven these days, and that can be an issue. Hopefully this example helps shed some light on what actually happens when immigrants move to the United States. 

Ohio economists say immigration counteracts Ohio’s brain drain

In a survey released this morning by Scioto Analysis, 17 of 19 economists agreed that an increase in the proportion of skilled immigrant workers has helped counteract human capital loss known as "brain drain" in Ohio over the past ten years. “For a long time, Midwestern states have lost college-educated labor to other regions.  In Ohio, at least, immigrants are somewhat more likely to have a bachelors degree (about 3 percentage points) and substantially more likely to have graduate degree (about 12 percentage points) according to American Community Survey data,” wrote Curtis Reynolds from Kent State.Wesleyan. 

Similarly, 14 of 19 economists agreed that international immigration has driven a significant portion of Ohio’s Gross State Product over the past 10 years. “Although only six percent of the workforce is foreign born they tend to be in high productivity sectors such as technology and health care,” wrote Robert Gitter from Ohio Wesleyan. 

When asked about the impacts that immigration had on low-skill wages, 11 of 19 economists disagreed that international immigration led to decreased wages. Seven more were uncertain about the impact, and only one respondent agreed that immigration resulted in lower wages. 

Will Georgic from Ohio Wesleyan wrote in his comment “Since the immigration is disproportionately high-skill, it seems plausible that recent immigration has applied upward pressure on low-skill wages in the state by increasing demand for goods/services to a greater extent than the competition for low-skill jobs increased.” 

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. 

Ohio officials turn down billions of dollars in federal aid

On April 7, Jake Zuckerman from Cleveland.com reported nearly 400,000 Ohioans have lost Medicaid coverage since pandemic-era enrollment rules were rescinded.

We all know the importance of health insurance coverage to a family. Illness or injury can plunge even a financially sound household into dire straits. But for those teetering on the edge of poverty or living in it, insurance coverage is even more important.

But there is another story to policymakers’ choices to restrict health insurance coverage for low-income residents: the lost economic development potential from federal Medicaid money left on the table.

According to the Kaiser Family Foundation, a leading source for health policy research, over two-thirds of Medicaid spending is picked up by the federal government. Kaiser also reports annual state Medicaid spending in Ohio averages over $6,900 per enrollee.

Assuming those who lost their Medicaid coverage are similar to the average enrollee, Ohio lost nearly $1.9 billion of federal dollars from denying Medicaid coverage for these Ohioans.

Let’s put that number in perspective. $1.9 billion is larger than the FY2024 budget for 57 of Ohio’s 61 state agencies, coming only behind the Departments of Medicaid, Education, Higher Education, and Rehabilitation and Correction. This is more money than Ohio spends on 44 different state agencies combined.

What does leaving $1.9 billion of federal funding on the table mean for the economy? It means less dollars going to hospitals and the greater health care system to treat Ohioans struggling with disease and disability. This means not only less resources for our health care system, but also less money in the pockets of nurses, doctors, and administrative staff to spend in the state economy.

This matters for a state like Ohio. Five of Ohio’s top ten employers are in the health industry including its top statewide employer, the Cleveland Clinic, which employs nearly 60,000 Ohioans. Health care is a $59 billion industry in Ohio, making health care 17% more concentrated in Ohio compared to the United States as a whole.

Much of this is driven by Ohio’s aging population. 18.3% of Ohioans are retirement age, higher than the national average and making Ohio older than 31 other states and the District of Columbia. This means Ohio has a greater need for a strong health care industry than the average state.

Beyond the direct economic effects of turning down federal funds for Ohio’s health care industry, there are also dynamic problems that come from reducing investment in health care. Losing out on incentives for preventative screenings can take small problems and make them worse, leading to chronic issues for workers that impact productivity.

Burdens from the cost of health care can cause people to skip out on medication or procedures which can help treat illnesses or conditions that end up hurting them in the workplace.

Legislators and other policymakers have made it clear that they care about Ohio’s workforce and economic vitality. Leaving billions of federal dollars on the table is not the way to do that.

This commentary first appeared in the Ohio Capital Journal.

Defining standing: whose benefits count?

One of the most important parts of cost-benefit analysis is clearly defining standing, or whose costs and benefits will be estimated in the analysis. In many ways, the problem of defining standing is similar to the problem of choosing which outcomes to quantify and monetize. In theory, we could keep thinking of progressively distant outcomes that may be influenced by our policy change. We could similarly ask which people are affected by our policy change further and further downstream. 

As an example, when looking at a minimum wage change in Ohio, we often focus on the impact of the policy on Ohioans. With infinite time and resources, we might be interested in modeling how an Ohio minimum wage could influence interstate commerce, but with limited resources, we might limit our standing to only those impacts on residents of Ohio. 

However, defining standing is not just an issue of deciding whose impacts we have the capacity to analyze, it often requires us to make judgements about what should or should not count in a cost-benefit analysis. One question that regularly comes up is whether or not we should give standing to impacts that are perceived as immoral or wrong. 

At last month’s Annual Conference of the Society for Benefit-Cost Analysis, I heard Richard Zerbe, one of SBCA’s founding members, talk about this issue. 

He suggested that it was sound guidance to only give standing to activities that are legal. If we accept that our laws are a reflection of the things our society values, it makes sense to let our laws dictate what has standing. Leave the questions of right and wrong to the lawyers and lawmakers. 

Of course, laws change over time, and there are countless examples of historical laws that certainly do not reflect the values of our modern society. This question became relevant to our work late last year, when we were working on our cost-benefit analysis of recreational marijuana in Ohio. 

One of the benefits we calculated was the consumer surplus that people would receive from participating in the recreational marijuana market. From an ECON 101 perspective, this is a critical part of the welfare equation. 

The question we were faced with was whether all of this consumer surplus was a new benefit, or if there was already  some consumer surplus that people who use recreational marijuana illegally were receiving. While it seems odd to only count that benefit assuming legalization occurs, it would go against the norm for cost-benefit analysis to include the illegal consumer surplus.

In the end, we ended up not counting the consumer surplus from the illegal market in our analysis. This was more a factor of the difficulty that came with estimating exactly what that would be, given the fact that we could never fully understand the size and prices that existed in an illegal market. 

I think if we were able to better estimate the illegal market for marijuana, this would have been a situation where giving standing to an illegal good would have been appropriate. Clearly, the legality of marijuana did not reflect its current social acceptance. This is evidenced by the fact that it is legal in almost half the country, and the vote passed quite convincingly in Ohio. As with every part of a cost-benefit analysis, the most important thing is to be extremely clear about what assumptions are being made.