Where Is the Next Puerto Rico?
C’mon guys. It’s New Jersey. Or… maybe Wyoming.
Puerto Rico is a fascinating demographic story because it has a unique combination of idiosyncratic migration, fertility, and mortality patterns, in a context with unique policy parameters, and an extremely high-profile bankruptcy ongoing. The thing is, if more people had paid attention to Puerto Rico’s trends, say, 15 years ago, some of the current collapse was entirely foreseeable just from the core population structure, and by the mid-2000s, the major migration trends were obvious in the available data.
That has led me to wonder: are there other examples of this? Are there time bombs out there that are ticking really, really loudly, that nobody is paying attention to?
To assess this, we need 3 things:
- An assessment of debt burdens or other liabilities, past, current, and future
- An assessment of demographic change, past, current, and future
- An assessment of economic productivity, past, current, and future
Doing this for one geography is a serious task. Doing a survey of numerous geographies is genuinely daunting. But I want to take a stab at it.
I’ll use a very simplified form of each of the above. For debt burdens and liabilities, I use Census’ State and Local Finance net debt from 2003–2015, and forecast out to 2020 basically based on the growth of net debt liabilities over the last 2–5 years. It’s a crude forecast, so undoubtedly has a huge range of error, but it’s an easy method that I can apply to many places all at once and at least get a not-entirely-stupid forecast. As an alternate measure, I use Pew Center for the States’ net pension liabilities from 2003 to 2015, forecast out the same way. Note that my net debt forecast for Puerto Rico reflects only publicly available financial statements and documents I found on Google, and is essentially a straight-line estimate based on the same computations I use for other areas: it does not reflect any knowledge or assumptions about any reforms that may or may not occur. It’s basically a placeholder, not an actual forecast of how the island’s finances will be addressed.
For demographic change, I use the 2010–2017 Census population estimates for states. I then conduct a structurally-founded, components-driven forecast of U.S. population out to 2020, from 2017. From there, I forecast out the change in each state’s share of U.S. population, based on the changes up to 2017. I then impute each state’s population by taking their share of U.S. annual population in each forecast year.
For productivity, I just take Gross State Product per capita, and extrapolate it forward based on average post-recession growth rates, with a few adjustments to deal with extreme values.
I’ll start with a simple question: which states have had or will have very high debts? I define very high debts as net debt greater than 7% of GSP, or net pension liabilities greater than 10% of GSP.
Let’s start with high net debt.
Oooooookay. So Puerto Rico is a league of its own. Then again, Puerto Rico has fiscal capacity other states don’t have. We’d really need to compare Puerto Rico to state and local debt plus net Federal debt. But that mixes funding sources, so from here on, I’m just gonna exclude Puerto Rico, and look at unusually-indebted states.
Much better!
You’ll note that most places have their lines sloping downwards. State net debt tends to rise in recessions, fall in expansions. My growth forecasts are benchmarked to post-recession growth, which means I am assuming economic expansion continues to 2020. Most economists give pretty decent odds on a recession before 2020, and I agree that the odds of that are good, but I wanted to show one very important thing: for many states, even if the current expansion continues, net debtswill never get to their pre-recession levels.
Take Nevada. Its net debt was under 5%. I don’t forecast it even reaching as low as 8% before 2020. I could absolutely be wrong! But at their current pace, certainly a return to 5% seems tricky.
At the same time, some states are showing steep increases in state debt even in the midst of expansion! Connecticut, South Carolina, and Maryland have all seen their debt keep rising well after the end of the recession. South Carolina’s finally fell in 2015, so I have it declining to 2020, but Connecticut and Maryland both rose or barely broke even in 2015. Now, they may decline some, but it took both states a really long time to crest the fiscal gap of the recession.
Now let’s look at net pension liabilities.
Many of these were just places that had a bad peak around the recession, and have since recovered, like Oklahoma. West Virginia is an interesting case too: very poor pension liabilities in 2003, but improvement since. The trick is WV had a pension system, then introduced defined contribution (DC) plans. The DC plans, however, cut off the number of payees into the pension, so the pension got horribly underfunded… finally, WV chickened out, and re-implemented a much-reformed pension plan, with bigger payments and smaller benefits. So now there’s a wedge of workers who will never claim benefits, and a growing crop of workers paying into a way-less-generous pension, and of course WV’s high mortality is thinning the ranks of beneficiaries at a despair-inducing clip.
Let’s restrict to just states with net pension liabilities over 10% in 2015 or later.
These are the states facing very-likely pension problems. New Jersey and Kentucky are fairly well-known. Kentucky because it is switching many workers over to a different retirement plan, depriving the existing plan of payees, New Jersey because… uh… I’ll admit I don’t totally understand. Basically… they promised generous pensions and then… didn’t pay for them, I guess? I’m not an expert on pensions. I’m sure there’s a more nuanced take.
Illinois also looks pretty dire… as does Mississippi. But hello Wyoming!
Wyoming’s problem is that the oil-and-coal bust has been very unkind to its economic output, and indeed to its demographic trajectory. While very low still as of 2015, if there’s no rebound in oil and coal, or if Wyoming doesn’t make a big funding change, its unfunded liabilities for pensions could rapidly escalate.
So I’ll score every state on these debt and liability factors by ranking their net debt as a percent of GSP forecast in 2020, the estimate in 2015, and the change between the two. I’ll also score every state by net liabilities in 2020, 2015, and their score. When I average up all these ranks, I can then correlate them with credit scores for state debt. I assign credit scores such that AAA rating is a score of “1” and each level down the credit rating scale is 1 integer higher.
Now, the upper right dot is Illinois, and having it in the frame makes it look like a pretty good relationship between these two variables. But actually, if you drop Illinois, it’s not that strong. With Illinois, we’ve got an R-squared value of 0.32. Without, it falls to 0.25, but the visual relationship is extremely uncompelling. The reality is that the index spread is quite wide for any given credit rating… and that a given index score can have very different credit ratings!
I give four states index scores of 40 or higher: Kentucky, New Jersey, Illinois, and Connecticut. Kentucky’s bond rating is AA+ as of 2017. New Jersey and Connecticut are tied for the second-worst bond rating grade of U.S. states at just AA- (only Illinois does worse than AA, and it languishes at BBB-).
Why is Illinois so much worse than Connecticut or New Jersey? Well, a likely factor seems like politics. Plus, Illinois has been more dependent on IOUs and has had more failures to pay current obligations. Illinois is also bound in a tighter constitutional straightjacket than Connecticut and New Jersey. That’s all perfectly rational, and could easily justify treating Illinois worse.
But while some of that is prospective, some of it is retrospective. Credit ratings should reflect expectations about the future. Will Connecticut be able to pay its dues? Well, As of 2015, Illinois’ state debt was 7.3% of GSP, and stable. Connecticut’s was 9.3%, and rising. Now granted, Illinois’ pension liabilities look a lot worse: 15.3% of GSP in 2015, vs. 10.8% in Connecticut. But on the other hand, New Jersey’s pension liabilities are a whopping 24.1% of state GSP, and rising! Mississippi’s are over 14%! And yet Mississippi is AA, New Jersey and Connecticut AA-.
Now, to be clear, I’m not making the argument that the rating of Illinois is unfair or wrong. I’m also not making the argument that the rating for these other high-index-value-good-credit-score places are wrong. But I am absolutely making the argument that the combination of high index values and good credit scores requires some exceptional justification. Why is Mississippi so creditworthy? It’s population has been in decline since 2014! Its GSP per capita is low, and, relative to the national average, has actually been falling since 2008.
That’s just one example. I could find others. South Carolina, Colorado, and Kentucky also seem to be getting pretty gentle treatment from bond-raters, while Wisconsin, Montana, and (surprising one!) Michigan seem to be getting stiffed a bit. There may be good reasons for these choices related to political commitments or even different assessments of future trends in underlying variables of course.
I’ll be returning to this topic more in the future, especially now that I’ve got some basic numbers to work with. And I’m sure #muniland people will take umbrage at one or another thing that I’ve said. That’s fine; fire away. My broader point is quite simple: there are a lot of sleepers out there. Another recession will throw dozens of states into Illinois- or New Jersey-level shortfalls on pensions, debt, and unpaid bills. And beyond that, the graying population alongside population shifts mean many of these obligations are becoming hard to sustain anyways.
I’m an an Advisor at Demographic Intelligence, the nation’s leading producer of rigorous national- and regional birth and marriage forecasts. I’m also a Research Fellow at the Institute for Family Studies, a Senior Contributor at The Federalist, and an Advisor at Demographic Intelligence, the nation’s leading producer of rigorous national- and regional birth and marriage forecasts. I’m a native of Wilmore, Kentucky, a graduate of Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth. I am not paid one penny by anybody for this blog post.
Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!
Check out my Podcast about the history of American migration.
DISCLAIMER: My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research.
