Comparing Apples to Apples: How Can We Measure City Budgets?
Last month I wrote about the fact that, unlike states, cities’ fiscal health seems to be worsening. New Census of Governments numbers show that local government tax collections declined 2.3 percent in 4th quarter 2010 compared to a year earlier, while state taxes jumped by almost 8 percent and total state and local tax collections climbed 2.2 percent. The dip in local taxes is largely driven by declining property taxes and, when coupled with declines in state aid, may force city officials to make tough spending choices.
Keep in mind, though, that revenue trends for local governments and cities are hard to read since the functions and services they provide vary more city to city than state to state. Some cities pay for schools and some don’t. Ditto for police, fire and myriad other public services. Differing service menus make it hard to decide whether a city is taxing its residents too much or too little. Over the years, researchers have used different metrics to compare cities by controlling for services provided.
Howard Chernick, Adam Langley, and Andrew Reschovsky introduced a concept of constructed city revenues at a conference we held last year. They apportion revenues, collected by local jurisdictions that include the city, based on how many of the jurisdiction’s residents live in the city itself. So a city that has two-thirds of a county’s population would count two thirds of the county’s revenues in its revenue base. Using that measure makes a big difference.
For example, 2008 Census of Governments data show that Baltimore per capita spending is four times that of Milwaukee, largely because Baltimore pays for school spending directly while Milwaukee has an independent school district. If this difference is taken into account using the Chernick-Langley-Reschovsky approach, it’s practically identical. Similarly, while Houston spent $1,625 per capita in 2008 and Providence spent $4,261, constructed city spending is about the same. The new approach lets us more systematically examine how different places are doing and begin to compare past spending levels and possible future shortfalls.
In a more recent paper, the same authors used constructed revenues to examine the relationship between house prices, state aid and forecasted city revenues, and spending. Based on those calculations, they forecast on average a 7- percent decline in city spending over the next three years. Some cities --mostly where housing prices tanked-- are expected to fare much worse than others. In the throes of the double trouble described by Kingsley and Lerman in earlier Metrotrends commentary, hard- hit Las Vegas, Reno, Riverside, Modesto and Stockton, for instance, may be forced to cut spending by 10-18 percent.
While the numbers in these forecasts seem too high to me, I applaud efforts to forecast city revenues and to provide meaningful measures of how cities are doing. They warn metros already facing the double whammy of declining employment and house prices of budget trouble ahead and, I hope, give them some more time to take action.