It’s a long way to fall if you can’t stick the landing.

What Are Puerto Rico’s Prospects?

What Were Ireland’s in 1845?

Puerto Rico has declared “a form of” bankruptcy, says the New York Times. This post won’t dip into the world of arguing about Puerto Rico’s core financial problems. But it will get into the question of Puerto Rico’s demographic future. This post is yet another update in a long series of posts I’ve had on Puerto Rico. It started here when Kristi Culpepper asked me to look at Puerto Rico’s migration record, and I showed truly historic levels of depopulation. Then a few months later I revisited, and showed that Puerto Rico was not primarily experiencing brain drain, but rather family drain, which is even more dangerous, because it siphons off the source of future growth. And then Salim Furth bugged me about looking at recent air traffic data for Puerto Rico, so I showed that, yes, migration really is that bad for Puerto Rico. Then last October, I updated ACS and airplane data to show that Puerto Rico might have hit rock bottom in 2015.

Well, we now have 2016 census estimates and new air traffic data! Huzzah! So let’s update, and then look at the future.


Air Passenger Balance Still Very Negative, But Improving

The chart below shows the net balance of air passenger traffic for all Puerto Rican airports, cumulative over the preceding 12 months, broken out by domestic (US) and international balances, going back to October 2002-September 2003.

I’ve included forecasts out to July 2017 for reasons I’ll get to in a bit. But, for now, it’s enough to see that although Puerto Rico’s 12-month net passenger balances remain deeply negative, they continue to be somewhat less deeply negative than recent lows, and domestic balances in particular are improving (though offset by worsening international balances). Thus, although Puerto Rico’s population hemmoraging will continue, it might slow down.

But wait! Could net balances be declining simply because population is falling?

Well, I don’t have monthly population estimates for Puerto Rico, but I can come up with some semi-plausible ones to compare to lagged 12-month averages. If we then compare these lagged net balances to my estimate of lagged 12-month population, we get:

This is basically the same chart, suggesting that Puerto Rico’s improving air traffic balances have improved faster than population has declined, which is a vital sign. If we’re looking for hopeful signs of population recovery, we don’t want nominal balances to become less negative just because there aren’t as many Puerto Ricans who could leave.

Now then, let’s compare cumulative years ending in July to Census migration estimates, which also reflect July-July population changes.


Comparing Migration Data Sources

The chart below shows Census Population Estimates net international migration for Puerto Rico alongside July cumulative net passenger balances for Puerto Rico. I use July because Census is estimating a July population finger, thus July-July population change. Of course, they key input of their survey is a March Survey asking about migration since the previous March… but oh well, let’s assume Census carried out any needed correctives to get up to July.

I’ve also included the ACS here, but I should say, there’s some weird stuff going on here. Census PEP migration data more closely mirrors ACS than it usually does for state-level entities, which is because Census relies on ACS for international migration, and they count all inflows/outflows of Puerto Rico as international. Tellingly, Census also lists a flat zero for its population residual field for Puerto Rico. In other words, they’re throwing their hands up in despair.

But here’s the weird quirk. ACS data is collected in monthly surveys from Jan-Dec of the target year (say 2015). But it asks about migration within the previous 12 months. So a January respondent is being asked about previous year (so, 2014) migration. We can weight each month of air traffic data by how many times that month’s migration showed up in ACS survey questions, then use the monthly air traffic balances and create a weighted sum to look at what net air passenger balance the ACS should have picked up.

And the weird thing is…it’s not even remotely the balance the ACS actually picked up. Simple July cumulative air traffic balance predicts the ACS (and PEP) data better than an ACS-configured air traffic estimate.

However, the July-configured air traffic balance, for whatever reason, does seem to be a fairly good predictor of PEP and ACS migration balances except in 2015, but 2015 remains recent, and I suspect Census may eventually revise it downwards. But if you trust my air traffic models as advance estimators of PEP/ACS, then you’d expect the 2016 and 2017 ACS estimates to come in showing migration roughly stable around -60,000, and you’d expect PEP to come in slightly higher, perhaps around -55,000 in 2017.

So if we plug all this into my long-run estimator of Puerto Rico net migration, what do we get?


Migration May be Stabilizing

Hey lookee there! Sure, total net migration is very low, but it kinda looks like it may be inching up! Of course, we saw periods of stability or even improvement in the mid-2000s as well, so stability for now doesn’t have to mean the true bottom in terms of outflows. And this is stability and record negative net migration, so, nothing to brag about.

Viewed as a percent of population, it looks like:

Ruh-roh. When I take all of my available sources and smooth them into one consistent time series back to 1910, it requires me to weight some sources, average across a few years, etc. And of course I’m inputing forecasts for 2016 ACS and 2017 ACS and PEP. But even though I forecast some improvement, my migration estimate as a rate doesn’t recover at all. It just kinda stagnates. Of course, this is not a record low rate of outflow; the 1950s saw more severe migration losses.

My expectation at present is that Puerto Rico’s net migration will remain roughly stable around -2%.

What Does This Mean?

Back in January, 2016 I made a range of population forecasts. Here is a demonstration of my estimates for 2015/2016 at the time, versus what Census ultimately estimated for 2016:

On the 2015 selection, you can see that Census did not make a large population revision for Puerto Rico.

But on the 2016 selection, you can see that Census’ population estimate for 2016 ultimately came in 15,000 people below my forecast, which I at the time thought fairly pessimistic. Whereas I had forecast a -1.38% growth rate, Census ultimately estimated a -1.78% rate. That’s pretty bad.

The above chart shows a range of possible population forecasts, where the only varying factor is what you assume about the net migration rate. This is similar to the forecast I made in my old post, but just updated with more recent year data. As you can see, for Puerto Rico to have long-run-stable population, they need positive net migration, because fertility is falling and mortality is rising. That’s a core headwind that Puerto Rico needs positive migration to overcome.

That positive migration, however, doesn’t come easily. Current net migration is about -1.5% to -2%, putting Puerto Rico on course for a population under 2 million people. I actually don’t know what modern economy to look at for a similar example of depopulation. Here’s indexed population of Puerto Rico vs. indexed population of Ireland, Ireland indexed and set at 1840, Puerto Rico at 2000.

Even in the context of modern population growth, Ireland hasn’t recovered its pre-famine population. Meanwhile, Puerto Rico’s population forecasts include within their range population losses even more severe than Irish losses.

We could be looking at centuries, or even permanent, depopulation of Puerto Rico.

Global population growth forecasts get lower and lower when we get out into the 2050s, 2060s, and later. Nobody makes forecasts out past 2100 with even a shred of credibility, but certainly nobody is out there forecasting a 22nd century population boom which could drive up Puerto Rico’s population as the 20th/21st century booms helped lift Ireland’s, via migration. There is fundamentally no reason to think that Puerto Rico has any more hopeful demographic future than “stable to negative,” barring some major policy shock.


What About Return Migration?

What about it? Here’s the share of Puerto Rican-identifying people residing in the mainland US moving to Puerto Rico each year in the ACS.

In recent years, return to Puerto Rico has been quite low. But maybe when conditions improve it will jump back up again. However, the long conditions are poor, the more likely it is that out-migrants put down roots on the mainland, and the less likely is return, especially as kids become school-aged.


Migration patterns can change quickly and dramatically, and Puerto Rico has had substantially less negative migration rates in its past. Thus, Puerto Rico’s population decline could reverse itself in the near future. However, at present, the best we can say is that outflows have stopped declining, and reaching stability at very negative levels.

If this level of outflow persists, Puerto Rico is facing a population collapse on the scale of Irish post-famine depopulation.

For those who think depopulation has big economic effects, we should really be studying Puerto Rico in great detail.

Check out my Podcast about the history of American migration.

If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!

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!

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.

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.