The left axis of the figure presents the number of people in Philadelphia County and the Philadelphia Suburbs (Bucks, Chester, Delaware and Montgomery) living below the poverty line (the red and green vertical bars). The right axis is a ratio of the total population in Philadelphia to the total population of the Suburbs (black line). In 1970 the total population of the two regions was the same with a ratio of 1. By 2005 the ratio had declined to just under .6 as Philadelphia shed half a million residents and the suburbs gained just under that amount.
The public discourse about economic development is often by necessity a kind of story telling about the causes and consequences of the state of the Philadelphia economy. A popular and compelling story is typically a version of the following from an article in the archive over at the Next Mayor:
“The answer to why Philadelphia tops the poverty chart is a depressingly familiar refrain about high taxes and a hostile culture that encourages affluent people to leave, while providing inexpensive housing that keeps poor people in place. "We were created when cities had a monopoly," said Jeremy Nowak of the Reinvestment Fund. "So we have monopolistic systems and viewpoints, when capital no longer has to be in any one place." Anachronistic regulations enforced by indifferent bureaucracies – remnants of the era when cities were the center of commercial life and had no competition - discourage businesses from coming here and creating jobs.”
So as Philadelphia lost population and employment the Suburbs gained both. And yet even in the Suburbs the number of people living in poverty (as in the City) has changed little in the last 35 years despite all that growth.
The point is not to suggest that attracting people to the City is unimportant but that policies aimed at doing so are often dressed up as though they will simultaneously reduce poverty. If you increased the population of the City by half a million, the poverty rate would fall from 25 to 18 percent. Would this outcome still count as a victory in the war on poverty if there were still more than three-hundred thousand people living in poverty? How much tax revenue in the form of tax breaks, abatements and incentives would you have to give up to buy back 500,000 people? And with what you can offer in terms of cash would they even come? Which is the bigger incentive a tax break/abatement or a community debating emergency measures to tackle a rising murder rate?
It is good policy to improve the competitive strengths of local businesses whether it is through investments in the workforce, improvements in city services and yes better tax policy. It is bad public policy to frame policy changes aimed at changing the competitiveness of business as solutions to poverty when these changes alone are plainly not sufficient.
The reason it is bad policy is because the cost of failure is so high. In 1970, just 11 percent of the people living below the poverty line were between the ages of 36 and 53. In that same year 44 percent of the people living below poverty were under the age of 17. By 2005, most of those kids captured in the 1970 census would be between the ages of 36 and 53, a group whose share of the population living in Poverty nearly doubled to 20 percent. As you ponder the next four years the most important question is not who will be the next mayor. It is what you can do to make sure that on your watch, whoever the next mayor is, another generation of kids do not have a similar fate?
*Note:
An error in my code misclassified some people in the 1970-2000 Census as living in poverty when in fact their poverty status is unknown in the data set. The error overstated the level of poverty from 1970 to 2000. At some stage I will impute poverty status for those that having missing data but for now I decided to exclude them entirely from the estimates for both regions. All the figures have been updated. In addition to revising down the poverty and population counts the changes also altered slightly the distribution of people by age and school enrollment. The points made above have not changed: The number of people living in poverty has not declined in the suburbs even as the population and employment have increased. The share of individuals age 36 to 53 in Philadelphia living below the poverty line has increased from 11 to 20 percent since 1970.











As usual
Another great post, Price.
When did it start?
When did the poverty start? Has there been a half million in poverty since the 50s? Since the Depression? Since before that? Was there a spike and decline from the Depression or just a plateau? Did poverty drop during WW2?
I realize that data may not be as available for my questions, but thought I'd throw it out there.
I'm also wondering if you could toss up a number difference chart rather than percent chart on the unemployment thread.
Thanks for providing data to the discussion
----
I support Michael Nutter for Mayor. My slate.
Not high hopes
Aardhart, you will notice above that I caught an error thanks to your question about earlier data so thank you. Probably not much more I can do back in time but I will look there is some data from 1950. I will try to post the unemployment numbers you want when I get a moment. Thanks again.
A start?
Not sure this captures your request but here are the monthly numbers on unemployment.
You are welcome for taking my critiques
Price, I'm glad my comments helped you out.
Thanks for posting the numbers. I was hoping for the second chart in this post except with a gross difference instead of a ratio. I think the ratio is just high when national unemployment is low and vice-versa. I don't think the ratio is that informative. (The gross difference might not be either, but a picture would help figure it out.)
----
I support Michael Nutter for Mayor. My slate.
question for Aardhart on the poverty thread
Aardhart wrote:
Not sure I’m following you. Lets start with what you mean by gross difference? Total Unemployment in Suburbs – Total Unemployment in Philadelphia?
Aardhart wrote:
Critiques? Perhaps you can elaborate further on the point embedded in your questions.
Price check
Lets start with what you mean by gross difference? Total Unemployment in Suburbs – Total Unemployment in Philadelphia?
Yep. I think the difference is a fairly steady ~3%. I think the chart of the ratio just rises when the unemployment is low (6/3=2) and falls when unemployment is high (12/9=1.3), but don't see the difference in ratio to be meaningful for analysis of the Philly region.
I can't tell really any trends and it is difficult to tell when the gap grows and shrinks, especially when it is steep. I'm not sure if the chart would really tell us anything, but the gap seems to grow in the 70s, shrink around 2000, and grow again. I don't know what that means, but I think the picture would help illustrate any significant time periods. Is Philly more susceptible to recession perhaps? Or maybe the chart would be meaningless.
Why did you chart the ratios instead of the difference?
Critiques? Perhaps you can elaborate further on the point embedded in your questions.
They weren't critiques. They were just questions. But I'm glad they helped you out. You are welcome.
I don't have a point yet. I'm skeptical of the notion that a locality can do nothing to lower unemployment within its borders (which grossly misstates your point), and I don't think that your charts prove that taxes have no effect on anything (which I don't think you claimed (but others have)). I don't think either of those are my points. As I've stated before, I think excessive taxes have had a negative influence on economic activity within Philadelphia and has contributed to many of its problems. I'm skeptical of any claim otherwise.
In any case, thanks for actually providing some numbers.
-----
I support Michael Nutter for mayor in November.
Here you go Aardhart
Aardhart wrote:
Ok now I see what you want.

Aardhart wrote:
Well in this example you get more or less the same answer with both.
On the other hand consider this example.
Here the unemployment rates fell in both County A and County B and as they did the difference between the two rates declined from 6 percent to 2 percent. But the unemployment rate in County A is still twice as high as the unemployment rate in County B despite the difference between the rates being just 2 percent (was 6 percent). So I choose the ratio because the question we were exploring was whether the Philadelphia unemployment rate had improved relative to the Suburban rate.
With the actual data for Philadelphia you can see this pattern in the 90s. As all unemployment rates fell the difference between the unemployment rates in Philadelphia and the Suburbs declined to a historic low of 2 percentage points in 1998.

However as you can see below the Philadelphia Unemployment rose from 1.4 times the Suburban unemployment rate in 1992 to 1.7 times just before the recession of 2001.
You are correct this data does not “prove that taxes have no effect on anything” but it illustrates like the opening of this thread just how outsized the public discourse over the economic development effects of tax policy is.
Thanks for posting the chart.
Price,
Thanks for posting the chart. To my eye, it seems like the point difference is more statistically significant (or statistically independent) than the ratio. (I don't know if that sentence makes sense or how to define statistically significant. However, I'm guessing that the inverse ratio (suburbs/philly instead of philly/suburbs) would have a much higher correlation coefficient (is that a term?) with the unemployment rates than the point difference would. Of course, random noise would independent, but meaningless.) The ratio doesn't seem to be as good a measure of the differences between Philly and the suburbs (and therefore of the city's policies). It seems like it just fluctuates with the regional and national trends (until about 2000). ()
The point difference trend does seem to show a downward trend until the dot-com bubble burst and 9/11. I'm sure that this trend started precisely in 1995 when tax reductions started. (Or maybe sometime in the 80s.)
I'm not sure what role taxes played in any of this. The point difference did decline under the early years of the tax reductions, but this trend did not continue into the new century. (Ratio did rise in the late 90s.) Maybe Street was viewed as bad by businesses or maybe it was factors outside of anyone's control.
I might agree with you that "the public discourse over the economic development effects of tax policy is" outsized. However, I do think that Philly has excessive taxes that do have a negative effect on its economy. (Even after 12 years of tax reductions, we are still maybe one of the highest taxed cities. Maybe that is why the effect is not that apparent.) I still favor tax reductions.
One advantage that taxes have over other factors is that tax rates are entirely within the city's power. They change with a law. The city can't pass legislation that will give 80% of the workforce a college degree.
Can you admit that the resistance to tax reduction on this blog greatly outsizes the negative effect that tax reductions have had on revenue. During 12 years of tax cuts, revenue increased. Yet, there is still talk of "blowing a hole in the budget," massive cuts to service, etc.
-----
I support Michael Nutter for mayor in November.
Aardhart, who was that masked man?
I don't think that your charts prove that taxes have no effect on anything (which I don't think you claimed (but others have)).
Did I miss it where someone said that local taxes have "no effect on anything," or where someone said "[his] charts prove that taxes have no effect on anything?"
One point that might be
One point that might be worth thinking about is whether the city or the suburbs do a better job of managing poverty -- or more broadly speaking, whether it's better to be below the poverty line in the suburbs or the city. (It may depend both on the suburb and on the city neighborhood.) I also wonder whether suburban poverty is less generational than urban poverty -- part of the promise of the suburbs, at least, is that the schools are better, the neighborhoods are safer, and there are in general more resources available to move out of poverty than there are in highly-concentrated poor neighborhoods in the cities.
Poverty as an absolute measure of income is one thing -- poverty as a relative measure of class and income mobility and quality of life is something else.
Interesting points, Short Schrift
Short Schrift I can’t speak intelligently to how the character of poverty in the Suburbs is different than in Philadelphia. I can however imagine reasons the difference might sometimes run the opposite direction you suggest. Access to public transportation and health care might be better in Philadelphia. With respect to the quality of schooling, if poverty in the suburbs is concentrated in particular school districts you get essentially the same outcome as in Philadelphia. On the other hand I would assume (may not be true) the likelihood of being a victim of a crime is greater in Philadelphia.
What is clear to me is that despite economic growth in the suburbs there hasn’t been a reduction in the absolute number of people living below poverty. So although starting out with a smaller number of people living in poverty and in theory more resources to tackle the problem the Suburbs would appear to have not done a “better job of managing poverty”.
I read these trends as meaning you have to have a plan to tackle the problem. You can't just hope the rising tide lifts all boats. If there is ever another federal effort to reduce poverty, given the budget situation it seems unlikely, it will probably require local financial contributions and most certainly local implementation.
Here are four figures distributing the share of people in poverty by age with the 18-35 age category split between those enrolled in school and those not. From this point of view there doesn’t appear to be dramatic differences in the two populations other than the poor Philadelphia population being somewhat younger.
And lastly here are two figures on the race and ethnicity of the two populations.