John O. Sullivan
“Anyone can love a perfect place, Loving Baltimore takes some resilience”
– Laura Lippman, American Fiction Author
For those of you who aren’t from Maryland, anything to do with Baltimore is a bit of a hot issue in state politics. Baltimore is a cultural and economic hub for the state, but even so it is a city plagued simultaneously by white flight, gentrification, economic depression, and crime. Baltimore receives more funding from the state government than any other jurisdiction, as such, many state Republicans see Baltimore as an example of everything wrong with expansive government and the Democrat-dominated General Assembly. So school funding in Baltimore was always going to be a hot political issue. When Baltimore City Public Schools was projected to have a $130 million budget shortfall, it became a very hot political issue. Add in a Republican governor (who favors charter and private schools) that the State Democrats would very much like to unseat, it becomes an extremely hot political issue. New mayor Catherine Pugh pledged to come up with a plan to solve the budget deficit. Her plan essentially consisted of asking the state for even more money, and that plan catapulted the school funding crisis from “extremely hot” to “literally on fire” as a state political issue.
One of the most prolific lines of attack leveled against city leaders by state Republicans in the midst of this crisis is that Baltimore City schools are a poorly managed disaster scene with not enough funding from the City and too much money from the State. In other words, the City should come up with the money to cover the budget gap, not the State. But despite all these calls for increased funding from the city or the state, it’s hard to know what exactly is the truth. So many different numbers fly around, and usually without context, so it can be hard to tell what any of it means. How much money is Baltimore spending on schools? Is it actually as low as claimed, or is it higher? If it is as low as claimed, is it still comparable to other cities?
To answer these questions, we did a (moderately) in-depth analysis. Baltimore is a city of roughly 620,00 residents (620,961 in 2010 census, 621,849 by 2015 estimates). All American cities with at least 500,000 residents but no more than 700,000 residents were examined as cities of comparable size. Washington D.C. was excluded, as the general premise of this article is to examine funding relationships between major cities and their states, and D.C. has no state. Memphis, Tennessee was also excluded, as its school system recently merged with Shelby County Schools and detailed funding data was not readily available. This created a sample size of 15 cities, including Baltimore.
The average population of this sample is roughly 630k, almost exactly the same size as Baltimore. The average growth rate of these cities is 6% per years (unweighted, I didn’t have all day), which is far higher than Baltimore’s. The only close comparison in the sample in terms of growth rate is Milwaukee, both Baltimore and Milwaukee showing under 1% population growth per 5 years. This is important because it can be a determiner in the overall economic health of a city. The median household income of this sample (also non-weighted) is $47,000, several thousand dollars above Baltimore’s $42,200. Baltimore is also one of the densest cities in the sample, second only to Boston, uncoincidentally an Amtrak Northeast Regional neighbor. Population density is an important determinant in finding “peer cities” because cities with different densities spend their money in different ways. High-cost, high-reward public transit tends to be much more developed in the dense metropolitan areas of the northeast and mid-Atlantic than in the flatter, less dense cities of the midwest (Oklahoma City was by far the least densely populated city with less than 1000 residents per square mile).
Upon close examination of a wide range of demographic data of the sample cities, including but not limited to: Population, growth rate, area, population density, racial makeup, Median household income, and age breakdown, two things became clear. 1) The cities were all unique. It was rare for any city in the sample to be a close comparison to Baltimore in more than a single demographic measure. 2) There was not yet enough data to form a conclusive pattern.
As a result of this, I, with the help of Tom, expanded the depth of the search into economic indicators beyond Median Household Income, that tried and true economic indicator, as well as expanding the breadth of education information available to us. These additional indicators included number of enrolled students, the estimated dollars spent on education per student, State population, percentage of all state residents living in the city, and median value of owner occupied homes. Surprise, surprise: Median Value of Owner Occupied Homes was the telling statistic. When all other factors were pointing every which way, Median Value of Owner Occupied Homes was a fair predictor of the level of state education funding a municipality would receive, with an r2 (correlation coefficient squared) score of 0.188. An R-score, also known as a correlation coefficient, is a measure for determining the degree of causality between two factors. An R-Score close to one means that the predicting factor determined almost all variation in the dependent variable. An R-Score close to 0 means that the variable determines almost none of the differences between data points.
This trendline is unsurprising because, as many of you probably already know, local school districts in the United States are largely funded by property taxes. Higher value properties result in more property tax revenue, which means more local funding for the school district. With more funding available from local sources, less is needed from the state budget to meet local school needs in the city, which results in the inverse trend of lower property values being correlated with higher levels of state funding.
This is a perfectly logical train of thought. So why such a low R2 score? An R2 score of 0.188 isn’t terrible in the social sciences, but it isn’t a holiday in St. Barts either. So, ever the Public Policy student, I dug even deeper. Here is the same graph, with a few of the cities removed:
As you can see above, this graph looks much more convincing. An R2 score of 0.774 is much, much, much better (Remember, R-scores are on a scale of -1 to 1). Which means the cities removed; Baltimore, Seattle, Las Vegas, and Albuquerque, are significant outliers. So this leads us to one final question: why do those four cities receive so much more state funding than their peers?
I have a working hypothesis that I believe explains the phenomenon: Then cities of Albuquerque and Las Vegas are by far the largest metropolitan areas in their states, both comprising over 20% of the population of their states. Seattle and Baltimore are closer to the average of the sample, both home to 10% of their state’s populations. But, what sets them apart is that they are both major port cities, two of the biggest ports in North America. My hypothesis is that the leaders of these cities, the mayors and state legislature representatives, leveraged their cities’ size and economically productive ports for favorable, even disproportionate, allowances in state education funding formulas and budgets. 21 of New Mexico’s 70 legislative districts are in the tightly packed Greater Albuquerque Region. 16 of the 42 members of the Nevada Assembly hail from Las Vegas (plus four from North Las Vegas). Those are impressive voting blocs which would wield significant political power. I can’t speak for Seattle, but I know from personal experience that Baltimore legislators, as well as the host of organized groups representing the various interests of the Port of Baltimore, are forces to be reckoned with in the state capital of Annapolis.
So, we come finally to our original question. Baltimore’s level of state aid for education is clearly higher than most other cities of comparable size. Does that mean that Baltimore City should find the money in its own budget to cover the education deficit?
I don’t know. That isn’t for me to decide for you.
Baltimore is a city with a lot of problems. In some ways, the rampant opioid crisis that is now gripping the country was born there (among other places) 40 years ago. Violent crime is a major problem. The city’s population is stagnating. Baltimore is experiencing a localized economic depression. I am not totally convinced that less state support is the right prescription for these ills. But I’m not going to use this space to talk about why I am skeptical of a Republican governor who favors charter schools telling a hugely Democratic city that they can find the money to fix their schools themselves. Governor Hogan certainly hasn’t done anything wrong, he was following the funding formula laid out by the legislature, a formula which, if this evidence is correct, would suggest that Baltimore legislatures have already gotten more than their fair share from the state. I will take this instead as an opportunity to reiterate how insane it is to fund local school systems through property taxes. It creates an inescapable cycle. Education is key to elevating the less privileged among us out of poverty, the key to revitalizing cities and communities. But when schools that are falling apart are half the cause of the problems, and the schools are funded from property taxes on an economically stagnant community, a vicious and inescapable circle is formed. When a community’s wealth is the primary determinant in the quality of its education, the poor are condemned to remain poor for generations.
Curious about the data we used in this article? You can check out my expansive spreadsheet here. Like this post? Want to get updates on new pieces, as well as occasional news and current events postings? Check out our Facebook Page for more.
Feature Image: Mayor of Baltimore Catherine Pugh left and Governor of Maryland Larry Hogan, right. Image Courtesy of CBS.