» With a budget stalemate in Congress, the future for transit funding may increasingly be in the hands of state and local governments. But that could magnify seriously inequitable outcomes, an analysis of data from 65 cities shows.
The federal transportation program is at a crossroads. Congress is apparently incapable of advancing new or expanded funding for roads and transit, and has even passed legislation cutting back on previously approved appropriations.
The stalemate has left academics and commentators grasping about for a solution. Some, as Eric Jaffe profiled in an article this week, suggest that a decline in Washington’s role in funding transportation infrastructure may lead to better decisions by states and localities about how to invest; too many projects, they argue, are poorly designed or executed, in part because of federal sway. In theory, states and cities will raise the funds for their transportation spending themselves and make better decisions as a result.
The federal government’s influence is undeniable, as illustrated by the vast expansion in streetcar projects around the country that followed the Obama Administration’s decision to pump money into them. While few critiques have mentioned that Washington’s role is far more oriented towards providing funding than mandating one approach over another, with the federal government pulling out, states and localities will be more likely to pursue their own course in terms of what kinds of transportation projects should be funded.
Missing from this conversation, however, is the arguably more important issue of how transit operations should be funded. While America’s infrastructure is deteriorating, the most important question for the large majority of the nation’s public transportation users (or potential users) is how often a bus will show up nearby and whether it will take them efficiently to employment or other destinations. While the federal government currently chips in for new vehicles, it rarely spends much of its own money on hiring drivers and paying for fuel (though it chipped in until the 1980s); most of those expenses are covered by passenger fares and subsidies from state and local governments.
But should Washington be more involved?
About a year ago, I reviewed data from 15 cities to evaluate how local funding affected outcomes. Because of the increasingly relevant public discussion on the issue of decentralizing transportation funding, I wanted to reexamine this issue with a larger dataset of all 65 U.S. metropolitan areas with populations of more than 800,000.* In order to conduct this analysis, I took advantage of data provided by the National Transit Database and the Brookings Institution. These sources provided comprehensive information about metropolitan area income and transit spending, both of which I have compared extensively below. Note that the charts below were generated using infogr.am and must be viewed on the website for interactive capability.
The findings, to summarize them quickly, are revealing of the significant potential downside of funding transit operations only at the local or state level. The data demonstrate that increasing local and state transit operations spending is closely correlated with metro area median household income and poverty rates. This is not the case for federal aid, as minimal as it is. In addition, though cities and states with more progressive electoral tendencies appear to be able to increase local funding for transit operations, that contribution may be significantly limited by the incomes of local inhabitants.
In other words, differing local incomes (the wealth of a region’s inhabitants) appear to be considerably affecting the level of transit service provided to the inhabitants of various regions. Since public transportation is a vital social service, this has the perverse impact of providing the least support to the regions that likely need it most.
This review suggests therefore that there is considerable reason to be skeptical of decentralizing transit funding. Indeed, it indicates that a centralization of spending at the federal level could improve outcomes in terms of regional equity by allowing a redistribution of resources based on need rather than ability to pay.
If local and state governments are tasked with assuming all funding responsibilities for paying for transit operations, the question is whether the result will be a significant divergence in funding for poor metropolitan areas as compared to wealthy ones (after all, it is worth remembering that American regions differ greatly in terms of income, which of course affects local tax revenues).
We already have considerable evidence, based on current local and state funding of transit operations, as shown in Figure 1.
The chart indicates that there is a relatively strong correlation between metropolitan area median household income and the combination of state and local per-capita transit operations funding (R-squared of 0.29). To a significant degree, this makes a lot of sense, since regions with higher levels of tax revenue are able to contribute a more significant amount of funding towards local services. Without aid from the national level, it becomes difficult for poorer cities to pay for the transit systems their wealthier peers can.
Figure 2 documents some of the very negative consequences of having income as a primary determinant of spending on transit operations. Regions with the highest levels of poverty (closely correlated in itself to lower levels of per capita household income) also clearly have the lowest levels of expenditure on transit operations per capita.
This is a paradox: The regions with relatively lower levels of poverty (such as Washington and Boston) can spend significantly more of their local and state funds on transit operations than regions with higher levels of poverty (such as Detroit and Memphis). None of this indicates that lower levels of poverty (or higher incomes) guarantee more transit funding, but rather that they allow for them. Poorer regions simply can’t afford to pay for better services.**
How can we explain why some regions pay for higher levels of transit spending and other do not? Figure 3 offers a suggestion.
Among the wealthiest metropolitan areas (those in the 4th quartile), there is a very strong correlation between the share of the vote President Obama received in 2012 in the core county of the region and the per-capita state and local spending on transit operations (an R-squared of 0.77). This indicates that wealthy regions with more progressive populations have pursued policies that result in higher spending on transit.
However, this appears to be less true for lower-income regions, even with progressive populations. Detroit, New Orleans, and St. Louis, for example, have very progressive populations in terms of their voting habits but do not spend nearly as much as other Democratic Party-voting cities. One clear explanation for the difference is that those cities have far lower median household incomes, reducing their tax bases.
State-level support for transit operations is also clearly influenced by partisan issues, which is good news for those who believe that states should decide independently what is “right” for them. Figure 4 compares the state vote share for President Obama in 2012 with the per-capita state-only spending on metropolitan area transit operations (not including local contributions), and there is a strong correlation (an R-squared of 0.44). There is a clear cutoff between states where a majority of citizens voted for the Democratic presidential nominee (many of which provided generous support for transit operations in their metropolitan areas), and states where a minority did (where most provided virtually no state support to metropolitan area transit operations).
This comparison raises a very significant problem with the idea of promoting a state and local monopoly over transit spending. Cities like Atlanta and St. Louis, which have considerable local support for progressive politics, are nevertheless handicapped by conservative-dominated state politics that refuse to contribute state funds to public transportation.
In sum, this evidence suggests that states and local governments, left to their own devices, will restrict funding on transit operations based on the income of their inhabitants, not based on need. It is not rational that the state and local funding for transit in San Jose is more than six times higher than that in Fresno, just 150 miles apart, much because of the latter’s significantly lower household incomes and more Republican voting tendencies. Fresno, after all, has more than double the poverty rate of San Jose and thus has a significant transit-dependent population that is not being appropriately served.
These data indicate that there is a potentially very important role for the federal government to act as a redistributive agent. If public transportation is an essential social service — almost as important to our society as Medicare or Medicaid or Social Security (that is what we think, right?) — then how is it fair for the people who live in the poorest metropolitan areas to suffer from inadequate transportation services? Why should cities, even if they have progressive populations themselves, have to suffer from low state funding because of conservative legislators?
At heart, this is the logic behind advocating a significant role for the federal government in funding public transportation operations. We cannot simply devolve decision making and financing to state and local governments and hope that the “right” choices get made, with no regard to the regressive impacts of, in effect, asking poorer cities to live within their means as the wealthier profit from theirs.
Of course, the federal government already does fund some public transportation around the country, and it sometimes, albeit not always, mandates rules. How well has this management functioned, in particular in addressing discrepancies between different cities based on their respective wealth?
Washington’s critics argue that the central government’s decisions have produced poor outcomes in capital construction. This may be true; the federal government’s interventions resulted in massive subsidies for automobiles over the past sixty years, with only a relative pittance devoted to transit. Why should we trust Congress or the Department of Transportation to make the right choices, based on that example?
But a clearheaded look at the evidence indicates that Washington has used its transit operations spending to advance redistributive principles. As shown in Figure 5, transit operations spending by the federal government in metropolitan areas across the country has no correlation with regional household incomes or poverty rates. The funding formulas developed by the Congress may not be perfect, but at least they are not discriminating between metro areas based on their respective incomes.
This indicates that Washington is, at least in this way, a reasonable custodian of government dollars when it comes to financing transit operations. To dismiss the federal government’s role is to ignore its important redistributive powers — its ability to transfer tax revenues from wealthier regions to poorer ones to help contribute to a more just society.
But the contribution of the federal government remains quite small, representing on average only about one-fifth of overall spending on transit operations in the U.S. Most federal transportation dollars are committed to capital expenditures, with just a small amount oriented towards operations. As a result, overall spending on public transit operations remains heavily tilted towards wealthier regions.
A more equitable funding system would take into account the differences in income and poverty rates across region and attempt to provide adequate transit services everywhere. To a certain degree, this is an impossibility in a country where so few national resources are devoted to running our buses and trains. Yet what seems evident is that a policy that orients funding towards states and localities will only encourage the unequal funding devoted to transit in rich and poor cities.
Update: It is worth questioning whether the differences in funding between high- and low-income metropolitan areas are, in part, a reflection of higher labor costs because of higher incomes there, as commentor Jeff argues. However, a comparison of metro area poverty and transit service hours (which are almost directly connected with transit service miles according to the data), shows very similar correlations as a comparison of metro area poverty and local and state transit funding, as shown in Figure 6 (the same is true for a comparison of metro household income with service hours). This reaffirms the argument that transit provision is to a significant degree the product of local wealth.
Note: Metropolitan area household income quartiles are defined as follows: 1st Quartile, $46,821; Median, $52,610; 3rd Quartile, $57,764.
* With the exclusion of Salt Lake City, for which inadequate information was available. The 65 metropolitan areas studied here are: Albany, Albuquerque, Allentown, Atlanta, Austin, Bakersfield, Baltimore, Baton Rouge, Birmingham, Boston, Bridgeport, Buffalo, Charlotte, Chicago, Cincinnati, Cleveland, Columbus, Dallas, Dayton, Denver, Detroit, El Paso, Fresno, Hartford, Honolulu, Houston, Indianapolis, Jacksonville, Kansas City, Las Vegas, Los Angeles, Louisville, Memphis, Miami, Milwaukee, Minneapolis, Nashville, New Haven, New Orleans, New York City, Oklahoma City, Omaha, Orlando, Oxnard, Philadelphia, Phoenix, Pittsburgh, Portland (Oregon), Providence, Raleigh, Richmond, Riverside, Rochester, Sacramento, San Antonio, San Diego, San Francisco, San Jose, Seattle, St. Louis, Tampa, Tucson, Tulsa, Virginia Beach, and Washington.
** That said, we must allow for the possibility of a chicken-and-egg problem; are the regions with better transit wealthier because of those transportation options?