CSY Replay #20: Advanced Statistics For Urbanists? Introducing The Population + Productivity Index

Sometimes population growth and economic growth move on different tracks.

The cover of the 1996 edition of the Places Rated Almanac. Source: amazon.com

(Note: Bill Fulton, the author of the wonderful newsletter The Future of Where, recently published really good piece that explores a point I’ve made in the past — people are often moving to places with lower economic growth rates, and leaving places with more economic activity. Bill’s focus was on dispersion to exurban areas from central cities and inner-ring suburbs, powered by the wider acceptance of remote work since the Covid pandemic. He argued that exurban areas may attract new residents, but it could be quite some time before they’re able to become true economic hubs or job centers.

Back in 2021 I came to a similar conclusion, but at the metro level. The data may have aged, but I think the underlying point still remains. Take a look at this piece below. -Pete)

Originally posted: September 7, 2021

I was always a fan of the Places Rated Almanac, even if the places I liked most rarely ranked highly in it. But it often gave great insight into all the factors that people consider when making living choices. Source: amazon.com

My recent opinion piece in Bloomberg generated quite a stir last week. In it, I show that population growth alone doesn’t tell the whole story of what happens in metro areas, and considering economic growth as well can change the perspective of metro success. In doing so I created a new measure I call the PPI – Population + Productivity Index. Here, I want to get into the data behind the article.

The PPI is simple: it’s the population growth rate plus the GDP per capita growth rate for a defined time period. If you’re wondering where the idea for this metric came from, look no further than the analytics community that developed advanced statistics for sports, particularly baseball, basketball and football. PPI is not so different from baseball’s OPS metric, or on-base plus slugging percentage. OPS might be the best statistic at determining who the game’s best hitters are: those who frequently get on base and are able to drive the ball for power. The hitters who can combine hitting frequency with hitting power are distinguished from the slap hitters who get on frequently but don’t hit for much power, or the free-swinging power hitters who don’t get hits often but can – and do – hit prodigious home runs.

The data I used was the 2010 U.S. Census and 2019 American Community Survey 1-year estimate, and the U.S. Bureau of Economic Analysis’ 2019 GDP for metro areas with more than 500,000 residents in 2010 (there were 106). I divided the GDP by the 2010 and 2019 populations. I found that if you average the population growth rates and GDP per capita growth rates among the 106 largest metros, on average population grew by 8.36% and GDP per capita grew by 32.29%. Using these figures the average metro area above 500,000 residents would have a PPI of 40.65%, or 8.36% + 32.29%. Looking at this, it’s clear that GDP per capita growth noticeably outpaces population growth. I’d have to do much more historical research to determine if that’s usually the case, but for the 2010-2019 period GDP per capita growth outpaced population growth at an average rate that was 3.86 times that of population growth. But greater weight is given by laypeople to the number of people drawn to a metro area, rather than its productivity.

Here’s four tables showing the data, split up into tiers to enable readers to more clearly see the rankings and rates. I’ll include discussion following each table. Please note that population and GDP growth rates for metros with above average rates have their numbers highlighted in green. Those with below average rated have their numbers highlighted in red. The first tier:

The first tier on the PPI scale are mostly high-performing metros that have been equally attracting residents and rapidly expanding their economies. All had population growth rates faster than the metro average of 8.36%. However, some attained their ranking on the strength of economic gains, sometimes greatly passing the GDP per capita growth rate of 32.29%. Of particular note are the top four metros on the scale – San Jose (80.30%), San Francisco (70.48%), Provo (50.48%) and Seattle (53.36%) – the only metros that had GDP per capita rates that grew faster than 50% over the period. But it’s also true that Austin (29.76%), Raleigh (23.02%) and Boise (21.49%) benefitted just as much from explosive population growth.

The second tier:

This group is slightly more variable in terms of the size of the metros and the growth path (either population or economic) taken. Miami, for example, had above average population growth and GDP per capita growth, 10.82% and 37.82% respectively. Meanwhile, Los Angeles had weak population growth (3.01%) but very strong GDP per capita growth (44.03%), as did New York City (1.69%/42.52%). But then we begin to see metros like Las Vegas and Jacksonville, which had very strong population growth but underperforming economic growth. Las Vegas gained 16.17% in population between 2010-2019, but its GDP per capita growth rate of 28.95% was significantly lower than the 106 metros average rate. Jacksonville performed similarly (15.90%/28.90%).

The third tier:

Here we begin to see metros that are under greater demographic or economic stress, or both. Chicago’s metro area population actually dropped by -0.03% between 2010-2019, yet its GDP per capita rate surpassed the 106 metros average rate and increased by 38.15%. Buffalo (-0.66%/36.34%) and Cleveland (-0.66%/34.14%) followed suit. On the opposite end were metros that added many people, didn’t grow their economies to the same extent. Colorado Springs (15.52%/20.83%), Palm Bay (10.78%/24.18%) and Oklahoma City (12.45%/20.23%) fell into that category.

The fourth tier:

The metros underperforming using this metric can be found here. Only three of the 26 listed (Indianapolis, Washington, D.C., and Durham, N.C.) added population at a faster rate than the metro average; only one of the 26 (Scranton, PA) had a GDP per capita growth rate that surpassed the metro average.

But that doesn’t tell a complete story. Next, I created a couple of four-quadrant tables and sorted the metros by the U.S. Census’ four general geographic regions for the nation (Northeast, Midwest, South, West) and also four subjective cultural/economic regions (East and West Coasts, Rust Belt, Sun Belt and Interior) to discern any regional patterns in the data. I color-coded the tables to better visualize the data for patterns. Well? Here’s the table by physical geography:

The picture tells a pretty strong story. All 20 metros identified as Northeast (sorry, Washington, D.C. is officially classified as in the South) were in the low population growth quadrants but were split nearly evenly (11-9) between the low and high GDP per capita growth rate quadrants. Midwest metros did slightly better; 13 out of 20 had low population growth, but 12 out of 20 had high GDP per capita growth. Metros out west had generally high population growth (16 out of 24) and high GDP per capita growth (17 out of 24). However, metros in the South performed quite differently. A full 25 of the 42 metros in the South had population growth rates that surpassed the metro average. Yet only 14 of the 42 had GDP per capita growth rates that surpassed the metro average.

How about using the cultural/economic regions I identified? First, let me say things change slightly because some metros have a physical geography that doesn’t match their current cultural or economic geography. So I took the liberty of moving some metros into other categories – Washington, D.C. is technically in the South but has increasingly become viewed as a coastal city; Pittsburgh, Rochester and Syracuse are in Northeastern states, but culturally/economically Rust Belt; Bakersfield and Fresno might be in the coastal state of California, but (in my uninformed opinion) seem more closely tied to metros in the nation’s interior. Here’s the table by cultural/economic geography:

The pattern strengthens using this approach. Metros on the east and west coasts are more clearly low population growth/high GDP per capita growth metros; so are Rust Belt metros. Metros in the Interior eked out a slight advantage in favor of high population growth/high GDP per capita growth metros. But Sun Belt metros were strongly high population growth/low GDP per capita growth in orientation.

What could this mean? I think it means that, despite their challenges with affordability and inequality, coastal metros have learned to become considerably more productive without the addition of great numbers of people, and Rust Belt and Interior metros are nearly done retooling their economies to follow the same path. We already know that deep inequality has become a feature of highly productive metros, and their viability is threatened without action.

Relatedly, I think it also means that Sun Belt population growth is masking the economic performance of many of its metros. It means that people may be attracted by Sun Belt affordability, open spaces and pleasant climate, but they may not necessarily be building the economic foundation needed to support their increasing needs. Perhaps the future challenge of many Sun Belt metros will be to create economies equal to the size of their metros.

In this context, Sun Belt metros remind me of actors who become stars on the big screen because of their good looks, or athletes that make in big in sports because of their elite athleticism. Their God-given abilities are what puts them in fantastic positions. But it’s what they do once there to continue excelling – actors honing their acting skills while taking on more challenging roles, or athletes deepening their knowledge of the game so they can win with their minds, not simply their speed or strength – that will set them apart.

Many will not take the next step. Unfortunately, many could suffer the same sad fate of athletes trying to beat their defenders long after the quick first step has disappeared, and actors trying to rely on good looks long after the beauty fades.