Data

Spending Like a…

Zachary Bartsch2 Comments

Is the federal government spending at a faster rate? Your answer probably has more to do with your biases than with anything else. Most people don’t know the numbers or they imagine some more appropriate past. Below is logged current federal expenditures (this does not include government fixed investment, only consumption. Yes, we can argue about measures. This doesn’t include transfers).

The line of best fit is about 1.6% per quarter or 6.4% per year. Golly! Our spending is rising so fast! But, US federal spending grew relatively slowly in the 90s – maybe due to that fiscal conservative, Bill Clinton. And our federal spending grew even more slowly between 2010 and 2016 – maybe due to that other fiscal conservative, Barack Obama.

But, inflation varied over this period. What about real, inflation adjusted federal spending? See Below.

Continue reading →

Why States Hate Nursing Homes

James Bailey3 Comments

Medicaid is a health insurance program for those with low incomes, funded largely by states. Overall it accounts for less than 20% of US medical spending. But there is one area where it is the dominant payer: nursing homes. Nursing homes are expensive, and Medicare (the typical insurance for those over 65) won’t cover them after the first hundred days, so most nursing home residents end up paying out of pocket until they burn through all their savings and wind up on Medicaid. At which point, Medicaid pays about $100,000 per year to the nursing home for the rest of their life.

States are responsible for up to half of that cost, and so start looking for ways to save money. One idea they have is to make it harder to build nursing homes: if there aren’t beds available, potential nursing home patients will have to stay home instead, where they can’t rack up Medicaid spending the same way. In fact, some states go all the way to a complete moratorium on new nursing homes:

Source: Institute for Justice

Some other states allow new nursing homes, but only with a special permission slip called a Certificate of Need (CON). CON is often required for other types of health facilities as well, like hospitals or dialysis centers. Research by me and others has generally found that CON doesn’t work as a way to reduce spending, and in fact actually increases it. CON might reduce the number of facilities, but that reduction of supply and competition gives the remaining facilities more power to raise prices.

So which effect dominates- does the smaller number of facilities reduce total spending, or do the higher prices increase it? It depends on the elasticity of demand:

In health care demand is typically quite inelastic, so the price effect dominates, and spending goes up:

But nursing homes could be an exception here. Elasticity of demand could be relatively high because of the number of potential substitutes- home care or assisted living for those with relatively low medical needs, hospitals for those with relatively high medical needs. Plus this is the one type of health care where Medicaid is the dominant payer. They could be especially resistant to price increases here, both due to their market power and their willingness to keep prices so low that facilities won’t take Medicaid patients (another way to save money!).

A new paper by Vitor Melo and Elijah Neilson finds that this is indeed the case. Indiana, Pennsylvania, and North Dakota repealed their nursing home CON requirements in the ’90s, and at least for IN and PA their Medicaid spending went way up. The paper uses a new “synthetic difference in difference” technique that seems appropriate, and creates figures that seem confusing at first but get a ton of information across:

They correctly note that they don’t evaluate the welfare effects of the policy; it’s possible that the extra nursing home beds following CON repeal bring huge benefits to seniors that are worth the higher spending. But nursing homes could be the exception to the general rule that CON fails to achieve the goals, like reduced spending, that advocates set for it.

Workers Finally Get a Real Annual Raise

Jeremy Horpedahl1 Comment

Back in December I pointed out that, thanks to slowing inflation, real wages had been rising since June 2022 (using either the CPI or the PCEPI for inflation adjustments).

With the latest monthly data, we can now report more good news for wage earners: CPI-adjusted wages have increased over the past 12 months. That had happened since 2021. In the past 12 months, wages of production and non-supervisory workers are up 5.1%, just a hair more than the annual increase in the CPI of 5.0%. It’s not much, and we’re not back to our pre-pandemic norm of 2% real wage growth. But it is more good news that we may finally getting past our post-COVID inflationary hangover.

GPT showing off on Conversations with Tyler

Joy Buchanan2 Comments

Since posting it in January of 2023, our post on a failure of ChatGPT has been one of the most popular articles of the year in terms of web traffic. Read ChatGPT Cites Economics Papers That Do Not Exist

LLMs will make up nonsense lies, but it might sound sophisticated and serious. As human editors, we are not used to detecting that kind of problem, because usually if something sounds serious that means it has been written by a serious thoughtful person (usually).

That’s a problem, but ChatGPT is really good at some tasks.

Instead of pointing out the fact that ChatGPT is not perfect, Tyler Cowen wants people to see the potential. He has done an entire podcast interview with ChatGPT, asking it to take on the voice and role of deceased writer Jonathan Swift.

I’ll share part of the transcript:

COWEN: In 1709, you, Jonathan Swift, still were calling yourself a Whig. What exactly induced you to switch to the Tory side of the ledger?

SWIFT: My shift from the Whigs to the Tories was a gradual and complex process, influenced by several factors, including personal connections, political developments, and ideological disagreements.

During my early years in politics, I was indeed more aligned with the Whigs, who championed principles such as constitutional government, religious toleration, and the supremacy of Parliament over the monarchy. However, over time, I became increasingly disillusioned with the Whigs, particularly as their internal factions and power struggles intensified.

The Whigs’ support for the war with France, which began in 1702, was a point of contention for me. I believed that the war was costly…

The replies seem nuanced and insightful. Nothing made me suspect that they were factually inaccurate, although I do not know enough about Irish history to judge.

Is there any human who could have produced this script? I think so, although it would have required a lot of work. If one of these replies is better than anything a human Swift scholar would produce, how would we know?

GPT4 can write good summaries for the work of a prolific author like Swift, because the model can train on lots of examples.

GPT4 could probably write a good biography of a modern figure by pulling together all of the writing by them and about them. Maybe GPT4 could efficiently scrape up all mentions of this figure online and synthesize them faster than a human scholar. However, we observed GPT3 completely making up citations when we tried to get it to do economics summaries.

I’m concerned that humans will use GPT4 to write but not do the requisite fact-checking. That could introduce a new corpus of work that the next LLMs will train on, which might be full of lies. Humans might not admit to using GPT, and therefore we wouldn’t have a mechanism for using extra scrutiny on AI-generated writing from 2023. Humans can make mistakes too… so the ultimate solution could be an all-powerful AI that somehow does begin with a fairly accurate map of the world and goes around fact-checking everything faster than human editors ever could.

Is Equity Crowdfunding Beating Adverse Selection?

James Bailey2 Comments

Most new businesses are funded with a combination of debt and the owners’ savings; equity funding has traditionally been relatively rare:

Source: Kauffman Foundation

Partly this has been a regulatory issue. Raising equity adds all sorts of legal burdens. Traditionally businesses could only accept equity investments from accredited investors and a small number of friends and family unless they did a full IPO and became public (hard enough that there are less that 5000 public companies in the US out of millions of businesses). This changed with the JOBS Act of 2012, which allowed small businesses to raise money from large numbers of non-accredited investors without having to register with the SEC.

Following the JOBS Act, equity crowdfunding sites like WeFunder emerged to match new businesses with potential investors. But equity crowdfunding has taken off relatively slowly:

Total dollar amount raised by regulated CF crowdfunding campaigns. Source: FAU Equity Crowdfunding Tracker

Its seen more success recently with some additional regulatory relief and the overall market boom of 2020-2021. But at ~$400 million/yr, its still well under 1% of all venture investment (~$300 billon/yr), which is itself tiny relative to the public stock market ($40 trillion market cap).

Why has equity crowdfunding been slow to take off? Partly its new and most people still don’t know about it. Partly early-stage companies aren’t a good way for most people to invest a significant fraction of their money; you probably want to be at least close to accredited investor levels (~$300k/yr income or $1 million liquid wealth) for it to make sense, and those at the accredited investor level already have other options. WeFunder is up front about the risks:

The other issue here is with asymmetric information and adverse selection. Its hard to find out much information about early-stage companies to know if they are a good investment; part of the point of the JOBS Act is that the companies don’t need to tell you much. The companies themselves have a better idea of how well they are doing, and the best ones might not bother with equity crowdfunding; they could probably raise more money with less hassle by going to venture funds or accredited angel investors.

I’ve long thought this adverse selection would be the killer issue, but my impression (not particularly well-informed and definitely not investment advice) is that there are now quality companies raising money this way, or at least companies that could easily raise money elsewhere. WeFunder has a whole page of Y-Combinator-backed companies raising money there. This week Substack, an established company that has already raised lots of venture funding, offered crowd equity and reached the $5 million limit of how much they could legally accept in a single day.

Overall I think this model is working well enough that I’m no longer in a hurry to become an accredited investor. Accredited investors have many more options for companies they can invest in and aren’t subject to the $2,200/yr limit on how much they can invest in early-stage companies. But even if I completed the backdoor process of getting accredited without being rich, I wouldn’t want to put more than $2,200/yr into early-stage companies until I was a millionaire, at which point I’d be accredited the usual way. And while most companies aren’t raising crowd equity, enough are that there seem to me to be no shortage of choices.

The Leading Causes of Death Among Elementary-Age Children

Jeremy Horpedahl7 Comments

You might have seen this chart recently. It comes from a letter published in the New England Journal of Medicine in April 2022. The data comes directly from the CDC. It shows the leading causes of death for children in the US. You will notice that firearm-related deaths have been rising for much of the past decade, and in 2020 eclipsed car accidents as the leading cause.

Many are sharing this chart in response to the recent elementary school shooting in Nashville. It’s natural to want to study these problems more in the wake of tragedies. After the Uvalde shooting last year, I tried to read as much as I could about the history of homicide and gun violence in the US, and to look at the research on what might work to reduce gun violence, which is summarized in a post I wrote last June.

That being said, I don’t think the chart above accurately characterizes the problem of elementary school shootings. It might accurately describe some broader problem, but it’s misleading with respect to the shooting we all just witnessed. The most important reason is that the definition of “children” here extends to 18- and 19-year-olds. Much of the gun-related homicides for “children” shown here are gang-related violence, not random school shootings at elementary schools. It’s not that we shouldn’t care about these deaths too — we very much should care — but the causes and solutions are entirely different from elementary school mass shootings.

Continue reading →

Deaf Census Speculations

Zachary Bartsch1 Comment

Between 1850 and 1910, most US censuses asked whether an individual was deaf. There were four alternative descriptions among the combinations of deafness and dumbness. Seems straightforward enough. The problem is that these aren’t discrete categories, they’re continuous. That is, one’s ability to hear can be zero, very good, bad, or just middling. What constitutes the threshold for deafness? In practice, it was the discretion of the enumerator. Understandably, there was a lot of variation in judgement from one enumerator to another. A lot of older people were categorized as deaf, even if they had some hearing loss.

Continue reading →

What is $100 from the Late Nineteenth Century Worth Today?

Jeremy Horpedahl1 Comment

Recently I was watching a lecture by historian Marcus Witcher which addressed the treatment of African Americans in the Jim Crow era. Witcher mentioned the “pig laws,” which were severe legal punishments given to Blacks in the South for what used to be petty crimes. Such as stealing a pig. He mentioned that the fines could be anywhere from $100 to $500, and then he asked me directly: how much is $100 adjusted for inflation today?

My initial, immediate answer was about $3,000. That turns out to be almost exactly correct for around 1880. But the more I thought about it, the more I realized that this wasn’t a satisfactory answer. We were trying to put $100 from a distant past year in context to understand how much of a burden this was for African Americans at the time. Does knowing that adjusted for inflation it’s about $3,000 give us much context?

Continue reading →

Life Expectancy By State 1990-2019

James Bailey3 Comments

I’m making a panel of historical life expectancy data by state available here:

Life Expectancy By State 1990-2019

It covers the years 1990 to 2019 for every US state, and has life expectancy at birth, age 25, and age 65. It includes breakdowns by sex and by race and ethnicity, though the race and ethnicity breakdowns aren’t available for every state and year.

This is one of those things that you’d think would be easy to find elsewhere, but isn’t. The CDC’s National Center for Health Statistics publishes state life expectancy data, but only makes it easily available back to 2018. The United States Mortality DataBase has state life expectancy data back to 1959, but makes it quite hard to use: it requires creating an account, uses opaque variable names, and puts the data for each state into a different spreadsheet, requiring users who want a state panel to merge 50 sheets. It also bans re-sharing the data, which is why the dataset I present here is based on IHME’s data instead.

The IHME data is much more user-friendly than the CDC or USMDB, but still has major issues. By including lots of extraneous information and arranging the data in an odd way, it has over 600,000 rows of data; covering 50 states over 30 years should only take about 1,500 rows, which is what I’ve cleaned and rearranged it to. IHME also never actually gives the most basic variable: life expectancy at birth by state. They only ever give separate life expectancies for men and women. I created overall life expectancy by state by averaging life expectancy for men and women. This gives people any easy number to use, but a simple average is not the ideal way to do this, since state populations aren’t exactly 50/50, particularly for 65 year olds. If you’re doing serious work on 65yo life expectancy you probably want to find a better way to do this, or just use the separate male/female variables. You might also consider sticking with the original IHME data (if its important to have population and all cause mortality by age, which I deleted as extraneous) or the United States Mortality DataBase (if you want pre-1990 data).

Overall though, my state life expectancy panel should provide a quick and easy option that works well for most people.

Here’s an example of what can be done with the data:

If states are on the red line, their life expectancy didn’t change from 1990 to 2019. If a state were below the red line, it would mean their life expectancy fell, which done did (some state names spill over the line, but the true data point is at the start of the name). The higher above the line a state is, the more the life expectancy increased from 1990 to 2019. So Oklahoma, Mississippi, West Virginia, Kentucky and North Dakota barely improved, gaining less than 1.5 years. On the other extreme Alaska, California, New York improved by more than 5 years; the biggest improvement was in DC, which gained a whopping 9.1 years of life expectancy over 30 years. My initial thought was that this was mainly driven by the changing racial composition of DC, but in fact it appears that the gains were broad based: black life expectancy rose from 65 to 72, while white life expectancy rose from 77 to 87.

You can find other improved datasets on my data page, and once again this life expectancy data is here: Life Expectancy By State 1990-2019

Regulatory Costs and Market Power

James Bailey1 Comment

That’s the title of a blockbuster new paper by Shikhar Singla. The headline finding is that increased regulatory costs are responsible for over 30% of the increase in market power in the US since the 1990’s. That’s a big deal, but not what I found most interesting.

One big advance is simply the data on regulation. If you want to measure the effect of regulation on different industries, you need to come up with a way to measure how regulated they are. The crude, simple old approach is to count how many pages of regulation apply to a broad industry. The big advance of Mercatus’ RegData was to use machine learning to identify which specific industry is being discussed near “restrictive words” in the Code of Federal Regulation that indicate a regulatory restriction is being imposed. But not all regulatory words (even restrictive ones) are created equal; some impose very costly restrictions, most impose less costly restrictions, and some are even deregulatory. Singla’s solution is to take the government’s estimates of regulatory costs and apply machine learning there:

This paper uses machine learning on regulatory documents to construct a novel dataset on compliance costs to examine the effect of regulations on market power. The dataset is comprehensive and consists of all significant regulations at the 6-digit NAICS level from 1970-2018. We find that regulatory costs have increased by $1 trillion during this period.

The government’s estimates of the costs are of course imperfect, but almost certainly add information over a word-count based approach. Both approaches agree that regulation has increased dramatically over time. How does this affect businesses? Here’s what’s highlighted in the abstract:

We document that an increase in regulatory costs results in lower (higher) sales, employment, markups, and profitability for small (large) firms. Regulation driven increase in con- centration is associated with lower elasticity of entry with respect to Tobin’s Q, lower productivity and investment after the late 1990s. We estimate that increased regulations can explain 31-37% of the rise in market power. Finally, we uncover the political economy of rulemaking. While large firms are opposed to regulations in general, they push for the passage of regulations that have an adverse impact on small firms

More from the paper:

an average small firm faces an average of $9,093 per employee in our sample period compared to $5,246 for a large firm

a 100% increase in regulatory costs leads to a 1.2%, 1.4% and 1.9% increase in the number of establishments, employees and wages, respectively, for large firms, whereas it leads to 1.4%, 1.5% and 1.6% decrease in the number of establishments, employees and wages, respectively for small firms when compared within the state-industry-time groups. Results on employees and wages provide evidence that an increase in regulatory costs creates a competitive advantage for large firms. Large firms get larger and small firms get smaller.

The fact that large firms benefit while small firms are harmed is what drives the increase in concentration and market power.

What I like and dislike most about this paper is the same thing: its a much better version of what Diana Thomas and I tried to do in our 2017 Journal of Regulatory Economics paper. We used RegData restriction counts to measure how regulation affected the number of establishments and employees by industry, and how this differed by firm size. I wish I had thought of using published regulatory cost measures like Singla does, but realistically even if I had the idea I wouldn’t have had the machine learning chops to execute it. The push to quantify what “micro” estimates mean for economy-wide measures is also excellent. I hope and expect to see this published soon in a top-5 economics journal.

HT: Adam Ozimek