The Federal Reserve Bank of Boston is out with a new study taking a close look at just why Rhode Island’s economy experienced such a severe downturn from 2007 to 2009 compared with its New England neighbors. The study highlights the state’s manufacturing mix and its housing bust as key causes.
The study was authored by Mary A. Burke, a senior economist at the Boston Fed. In a phone interview Tuesday, Burke talked with WPRI.com about what her research revealed about the Rhode Island economy and what the state should do next. The transcript has been lightly edited for length and clarity.
What made you decide to research Rhode Island’s economic struggles?
It’s something that has just been a question that just kept coming up, year after year. I periodically give talks on the regional economy, and we would sometimes throw out hypotheses. We’d say, “Oh, why did Rhode Island do worse?” And we would throw out these casual hypotheses. I realized over time that we never really had a rigorous answer to these questions – maybe we’d come up with our little theories, but we’d never really tested any of them.
Then there was a movement in our department to do more regionally oriented research that was really aimed at economic policy, using regional outcomes to help us understand larger economic questions. One of the questions that kept getting brought up as one we might look at is, why did Rhode Island have a worse recession? It was just always posited as, well, this would be a great question to look at, and I was, like, well, OK – I’ll look at it.
It was the kind of thing where we thought, “Well, don’t we know that?” And I said, “Well no, we don’t really know the answer.” It hadn’t really been looked at very carefully, and I would say that I don’t even feel now that I have a complete answer. I have a bunch of clues, but I still have a lot of unanswered questions.
I also know you went to Brown University. Did that help encourage you to do this at all?
Oh, yeah – I really had a wonderful time there. I loved Providence when I was there. I love going back to visit. I sometimes feel guilty that I didn’t stay – it’s the classic “brain drain” kind of thing. But I at least feel good that I’m close by and I care about the Rhode Island economy, and I think that does augment my concern.
So let’s go into your research. Your paper is academic, full of charts, and contains some complicated language for the average person to understand. How would you boil down your findings and the conclusions you reached in this paper about Rhode Island’s economic woes?
First of all, there was not one single factor that you could point to and say, OK, that’s it, that’s the explanation. There’s not really a neat, compact explanation. It seems almost that it was a perfect storm – there were many, many weaknesses in the Rhode Island economy and factors leading up to the recession that all contributed.
But I think the two biggest ones – and you pointed them out in your article – was [first] the weakness of the manufacturing sector in Rhode Island. That was a structural pre-existing condition that I think was exacerbated by the fact that we then had this national demand shock – a weakness in demand for goods in general – that hit Rhode Island’s manufacturing sector particularly hard because they had pre-existing vulnerabilities in terms of value added and competitiveness. So that was one important factor.
Another was the fact that Rhode Island had a very large housing boom prior to the recession, and then that led to its having a greater housing bust. This has been shown to be true: there’s a strong association at the national level between the extent of the housing boom in, say, a county or an area prior to the recession and how severe the recession was in that place.
You highlighted an academic paper I’ve also found noteworthy, the 2011 David Autor one about which parts of the U.S. were most exposed to competition from Chinese manufacturing imports. Providence was ranked No. 2 out of more than 700 zones for exposure – very high – yet it hasn’t been discussed much here. Why was Providence so exposed, and how important do you think that is to the story here?
That’s a really important question to explain carefully.
Part of it is just historical accident. Different places tend to concentrate in different types of manufacturing. You get a cluster of a specific type of manufacturing, because it helps people to be close to other people producing the same type of thing.
Unfortunately, this leads to a lack of diversification, so if you happen to be producing something that is then subject to import competition – such as jewelry – then your entire industry can get wiped out because you have a lot of jewelry manufacturers. I know that’s not exactly true – the entire industry didn’t get wiped out – it’s just an illustrative example. But the point is, not every place is perfectly diversified in the composition of its manufacturing sector, and like I said, a lot of that is historical accident.
Then you get what we call agglomeration economies. Once you have established a strength in a particular kind of manufacturing, you tend to get a lot of that type of manufacturing in the same place, which can be good while it’s going up and up; there was a time where Rhode Island was I think manufacturing 70% to 80% of all the jewelry in the United States, and so that was great when it was strong and when it was competitive. But then the ground changed, and it became less competitive, and so then you had this huge concentration in something that was vulnerable.
So that was the structural story leading up to the recession. But I will say that there’s a limit to how much that structural story can explain Rhode Island’s worse performance during the recession, and here’s why. If you look at the relative performance – let’s just say, employment growth – manufacturing employment growth was negative in all the New England states between 2000 and 2007. It was cumulatively the worst in Rhode Island, so Rhode Island had the largest percentage employment loss in manufacturing between 2000 and 2007 of any of the New England states – but not by a lot. Not by a huge amount. It really only overtook the other states in terms of that negative growth as of 2006 and 2007.
But then, if you look at the relative performance – the employment losses in manufacturing in Rhode Island during the recession, compared with the manufacturing employment losses in other New England states during the recession – suddenly the relative performance [of Rhode Island] becomes much more extreme. If you just took the ratio of the job loss percentage in Rhode Island over the next-worst state, it suddenly just gets larger.
So you can’t just say, OK, well Rhode Island was doing twice as badly or it was doing 20% worse than the other states going into the recession, and then it continued to do just that much worse during the recession – no, that’s not the case. This negative pattern, this worse performance, was exacerbated.
One of the unanswered questions is, I’m not sure exactly how this susceptibility to increasing import competition interacted with, then, the domestic shock of the recession. So it could be that, OK, [Rhode Island] had these manufacturing industries that were clearly weak and susceptible and not competitive with respect to foreign competition – that means that would have left them also susceptible to a negative demand shock, where that would mean people are just willing to pay less for stuff. People have less money, they feel less wealthy, there’s this negative demand shock, so anything that says that an industry is vulnerable – maybe the value added, maybe the profit margins over the costs are small – and then suddenly the price that people are willing to pay for something goes down, that puts the profit margin maybe into negative territory all of a sudden. It changes the calculus of what is profitable.
So like I said, the prior weakness suggests that already the profitability was low or marginal, and then this domestic demand shock could have made that worse. But I can’t fully explain quantitatively why Rhode Island did that much worse in manufacturing during the recession. The Autor paper explains why it did a certain amount worse prior, but I also don’t know what happened to exposure to import competition during the recession, because they only went up to 2007. And it turns out that the method they used to measure that increase in import exposure is difficult to measure during the recession because they’re using this exogenous shock. My guess is that import exposure might even have gone down during the recession because there was basically a world recession, so the amount of competing products from China might have actually gone down.
So I don’t think you could get that much bang out of that explanation, if you looked at the Autor measures during the recession. I think it’s just they were weak industries, and now they were exposed to another shock – the domestic demand shock.
Now, if you look at Exhibit 4A in the paper:
So here we do this exercise where you’re predicting what would have happened to employment growth in a state based on its overall industrial competition – not just in manufacturing, but overall. In other words, how much manufacturing did Rhode Island have versus how much retail versus how much education and health.
The surprising thing to me that came out of this exhibit is that, if you look at the prediction in the first row, which says “07-08 Bartik Shock,” that’s basically what I’m calling the predicted employment growth. It’s based on just saying there was this national demand shock and let’s just apply that to the sectoral shares – so if you happened to have a lot of a sector that had a big negative demand shock at the national level, you would expect that just projecting that into the local economy it would also have a big impact if you had a lot of that bad sector.
If you look at these predictions, they are not always worse for Rhode Island, and they don’t vary that much across the New England states. Based on the overall industrial composition, New Hampshire should have had the worst recession. So in other words, these Bartik shock predictions are always largest, and I mean in the negative sense, for New Hampshire than in any other New England state. But New Hampshire didn’t even have the second-worst recession. So another puzzle we need to explain in some sense is, why did New Hampshire do better than expected? And even some of the other states, Massachusetts actually did better than we would have expected based on this overall composition.
Now the simple answer is that just having a large manufacturing sector when manufacturing gets hit hard nationally doesn’t mean you will get hit hard, because everyone’s manufacturing sector is different. The lesson is, you have to drill down deeper and deeper in each sector and say, what kinds of manufacturing did they have? And that’s where then you go back and you get a little bit more explanatory power out of that, and I think that’s partly consistent with this Autor story. But I haven’t looked at the same level of detail that they have at, say, the national employment shocks to manufacturing at a very detailed industry level.
So I think if I were to look at this more, I’d like to drill down even more to more detail within manufacturing and say, well, which types of manufacturing were hit hardest at the national level in the recession, and did Rhode Island have more of of those? I think that answer is probably yes.
And if you look at Exhibit 9:
Here we have what we call the Bartik manufacturing shock, which is at the level of within manufacturing, now we predict what the job losses would have been just within manufacturing alone based on composition within manufacturing. And here we get a little bit of headway, because if you look at the first set of numbers for 2008, Rhode Island has the largest negative value for the Bartik manufacturing shock. But it’s not that much worse.
It’s very close to New Hampshire.
Exactly. And now you look at 2009 –
New Hampshire is worse.
Yes. So here’s where we did this regression, where we used these two factors to explain overall employment losses, and we did this for all 50 states. You get a large amount of explanatory power – you explain about 60% to 70% of all job losses – just based on this predicted manufacturing sector losses and the lagged house-price change.
And here’s where the house prices put Rhode Island over the top.
Wow, look at those numbers – nearly a 10% drop in house prices the prior year for 2009.
Yes. So the lagged CoreLogic house-price change was way worse in Rhode Island than in any other state: -5.8% for 2008, -9.8% for 2009, and -9% in 2010. And that’s when job losses were steepest, between 2008 and 2009. This is why we [measured] them lagged [by one year], though, because you have to worry that job losses can actually cause houes prices to go down indirectly.
The lagged house prices put Rhode Island in last place [in New England]. So in some sense Rhode Island’s rank in the house-price decline was more decisive than its rank in the manufacturing shock, because it wasn’t always last in the manufacturing shock. If you just looked at manufacturing, that wouldn’t have been enough to put it in last place. You need these higher house-price declines.
Another factor that gets discussed frequently down here is the tax and regulatory environment. As soon as I posted my story about your paper, I had people saying, “Well, why didn’t she look at taxes?” What’s the answer?
Here’s why – because they didn’t get worse. Because they didn’t change.
I haven’t looked at this in detail, but I would say that’s sort of like the background noise. It’s something I call a “fixed effect.” If that were the case, why did Rhode Island have a pretty competitive unemployment rate and employment growth rate in, say, 2005? Right? I don’t think that the tax and regulatory environments were so great in 2004 and 2005, and then suddenly they started raising taxes and making regulations worse in 2006 and 2007.
I’m just guessing, but I don’t think that that’s the case, because people have been complaining about the regulatory and tax environment for years. I need to check this fact, but I don’t think there was some precipitous change or decline in the business climate right before the recession. That would have had to happen, because Rhode Island didn’t look that much worse than the other states before the recession. In fact in some ways it looked better, because it had higher growth in house prices – some people would say because the regulations were too weak; it was too easy to get a mortgage, it was too easy to become a Realtor or a mortgage broker and set up shop, and it was loose regulations. So I’ve heard both sides of the coin on that story.
Now that of course doesn’t necessarily mean that the business climate is ideal or that it isn’t a problem, right? You’re just trying to explain what precipitated such a worse recession for Rhode Island from 2007 to 2009.
Yes. It could be that that interacts with it, but, like I said, a couple things don’t agree with that, which is Rhode Island looked pretty close to the other New England states before the recession, when presumably it had some of these worse business-climate factors in place. And as I pointed out, in the recovery period, job growth has been faster in Rhode Island than in some of the other New England states – faster than New Hampshire, which is touted for its more favorable business climate. So it just does not agree with all the facts of pre-recession and post-recession employment growth.
You just alluded to it there but I wanted to mention explicitly, your paper tries to turn around the conventional wisdom a bit about the idea that Rhode Island’s recovery is so weak. What you’re saying is, no, it just looks weak because the recession was so bad, but the recovery has more or less kept pace with New England.
Yes. If you fell farther, it’s going to take you longer to dig out of that hole, even if you’re growing at the same pace – you put it well in your piece, you said it would have to have grown much faster to have made up as much ground in the same amount of time.
Now, I think if you look at just Providence it’s been among the slowest for employment growth. But Rhode Island as a whole, if you look at the state level, it has not typically been the laggard in the region for employment growth. This is payroll employment growth; you also have to look at household employment growth.
But if you also look at how quickly its unemployment rate has fallen, Rhode Island’s unemployment rate has been falling much faster than the unemployment rates of some of the other states – partly because it was so high, but it has been coming down a lot faster, and employment growth has been in the middle for Rhode Island in the past couple of years.
It’s having decent, not gangbusters but decent employment growth. But it just had a deeper hole to dig out of.
So after hearing all this, the natural question people will have for you is, what should Rhode Island do? How should Rhode Island dig itself out of this economic hole? Or is that beyond the scope of this paper?
That really was beyond the scope, but one thing I did notice was that a lot of the similar-sized cities, when I looked at the NECTA level, I noticed that, say, Worcester and Springfield seem to be doing better than I expected. People say, “Oh, these are these dying industrial cities, they’ve lost all these manufacturing jobs, they’re never going to get them back.” Well, they didn’t get the manufacturing jobs back, but they got other jobs. They got education and health-care sector jobs, and those are the sectors that are growing nationally based on demographic trends.
So the idea is that hopefully Rhode Island should be able to capitalize off this wave, the structural reasons why we’re seeing growth in health-care sector jobs and education. It seems like Rhode Island should try to capitalize on those. Just the same way that it was the victim of these structural trends going against manufacturing, now it needs to try to ride the wave of the structural growth in these other sectors. The fact that we do see growth in these sectors in other, similar cities that lost manufacturing jobs means that it can happen in Rhode Island.
Again, I don’t know the particulars of what were the regulatory factors or what were the other advantages that Worcester and Springfield might have had – one was UMass Medical, I know in Worcester – but why can’t something like that happen in Rhode Island? I think there’s enough growth in the health-care sector, and those are services that need to be delivered locally – there’s obviously some regionalization there, but it seems like there’s room for growth there because it’s also less susceptible to import competition. Those jobs are recession-proof, practically. If you look at the education and health-care sectors, they did not lose jobs in the recession. So it would be nice to go after recession-proof jobs, and it seems like there’s room for growth.
Hospitality and leisure is also doing well in the region, and that’s doing well in Rhode Island, as well. So it needs to make sure it hangs onto that growth. It should look at, what is doing well right now? Where is the job growth now? Can you try to protect those jobs? And where is job growth in similar types of cities, and can they capitalize on those kinds of trends?
My only counter on that is we see such a push nationally and in Rhode Island to reduce costs in health care and education, so whenever people suggest it should be our employment growth engine I wonder, how do you square that with the desire to bend the cost curve and reduce the cost of college and such?
The cost curve is really about technology and the use of technology. You’re going to need staff. Everybody talks about nursing shortages and an aging population, a shortage of staff. They need staffing. The main costs are not as much the staffing as it’s the technology and the number of procedures that you’re doing. I think there’s still room for employment growth and cost containment, especially people like the nurses. The nurses aren’t super low-pay – those can be very decent-paying jobs – but like I’m saying, the big money is coming from the procedures, the very intensive end-of-life costs, all the investment in expensive technology, and the remuneration for those. And that goes to a combination of the doctors and the hospitals – some of it may trickle down to the nurses, but I think most of that cost is coming from technology.
Any final thoughts on Rhode Island and your research to close the conversation here?
I just think that maybe the outlook is better than what we might expect based on the experience of the recession.
Obviously, policy will matter going forward, so I’m not going to say that people who say the business climate could be improved are wrong. I’m sure there are things that could be looked at there. But I just think that there are also some strengths that could be capitalized on, and focusing on what it’s been doing right is probably the way to go instead of just saying, oh, it’s all bad, we’re just doomed – try to highlight those things that are going well, and try to get the lessons from those things.