What's a budget supposed to do, anyway?
/Last month, Philadelphia City Council and Mayor Jim Kenney struck a deal to pass the City of Philadelphia’s $5.3 billion budget for FY2022. With Philadelphia on pace for its highest number of homicides ever in 2021, much of the media focus was centered on the $155 million allocated for violence prevention, including a growing consensus for the importance of evaluating the public’s investment in violence prevention programs. Unfortunately, there was no discussion of a fundamental flaw in the way the City of Philadelphia presents its operating budget to the public: the budget doesn’t actually explain how much it costs to deliver a single city service.
But, isn’t that what budgets are supposed to do? It’ll be a little wonky, but I can explain why the City’s budget doesn’t do the thing you think it’s supposed to do. I’ll start by asking a question: how much does it cost to operate the Free Library of Philadelphia? As advocates called for increased funding for the 54-branch system last month, the Free Library’s FY2021 budget was cited as $39.3 million and it’s FY2020 budget was indicated to be $45.7 million. Are these numbers correct? The first place to look is the Finance Department website, where anyone can access a wide range of budget and financial reports. I looked for the ‘Budget in Brief,’ which summarizes the City's annual operating budget. The most recently available Budget in Brief was for the FY2020 Adopted Budget, which is budget nerd speak for the City Council approved operating budget for July 1, 2019 to June 30, 2020. During that fiscal year, the Free Library’s General Fund budget was $45.8 million, matching the number cited earlier. I was surprised, however, that the Free Library’s budget was presented with just three simple categories: Personal Services (the cost of people), Purchase of Services (the cost of contracts and services), and Materials, Supplies, and Equipment (the cost of stuff).
For greater detail on what is included in these categories, I turned to the Operating Budget Detail, a budget nerd's dream with so much information that it is split into two 1,200-page tombs. With the assistance of CTRL+F, I found the same $45.8 million for the Free Library’s budget, but I found something else as well: a line for Employee Benefits with no information. It turns out that Personal Services only includes salaries and wages, while the cost of medical insurance, pension contributions, and other types of non-cash compensation commonly referred to as "fringe benefits" for current employees fall under Employee Benefits.
As a public sector consultant looking at city budgets across the country, I know other big cities – New York, Chicago, and Houston – all include fringe benefits in their departmental budgets. Some cities, like Seattle, even include indirect operating costs for things like rent and utilities. The combination of these direct and indirect costs provides a complete picture of what it actually costs to run a department and/or deliver a service. Why don’t we do the same? There is a straightforward reason: fringe benefits are managed centrally through the Finance Department, so the $1.4 billion cost for these benefits citywide is housed within the Finance Department rather than with individual departments. As a result, it is unclear how 28 percent of the City’s budget is spent.
It makes sense for significant benefits like healthcare to be managed centrally, especially because the City is a self-insured organization that assumes the financial risk of paying for health benefits. A centralized system can be a good management practice with more robust controls by subject matter experts and strong accountability. It is also true that the City budgets using large expenditure classes, which combine multiple related costs and provide management flexibility, while many local and state governments use line-item budgets, which tend to focus on control. One significant benefit of a line-item budget, however, is that it provides a detailed breakdown of the specific costs for operating a department that delivers services to residents (see Pittsburgh’s police budget example below).
While there are debatable strengths and weaknesses for how the City manages its budget in practice, there is no debate that the City’s budget doesn’t accurately reflect the cost of operating any City department. Let’s look at two charts that will help illustrate the problem.
First, we have a pie chart that shows the City’s FY20 $5 billion General Fund budget. More than 40 percent of the budget is consumed by three slices of the pie: Pensions ($749 million or 15 percent), Police ($741 million or 15 percent), and Other Employee Benefits ($663 million or 13 percent). Looking at this chart, it would make sense to believe that running the Police Department costs $741 million and represents 15 percent of the General Fund budget.
The second chart shows the same FY20 General Fund budget by expenditure type, and the picture is different. It shows that total Personnel Costs are $3.2 billion, or 64% of the total operating budget. The $1.4 billion cost of Pensions and Other Employee Benefits is the same here as in the previous chart and matches the amount that is housed in the Finance Department’s budget. The main difference is that it consolidates the salary costs that were spread across departments in the first chart, showing a total of $1.8 billion citywide. As a result, we can see that the citywide total of $1.4 billion for benefits is equivalent to 76 percent of citywide salary and wage costs.
Let’s go back to the Free Library’s $45.8 million budget. Most of the budget is Personal Services – salaries and wages for employees – at $40.6 million. There is no budget or financial document on the City’s website where you can find the full cost of Employee Benefits for the Free Library or any other department,[1] but we can guestimate. Using the citywide ratio that the cost of benefits is equivalent to 76 percent of the cost of salaries, the Free Library’s benefits costs would be $30.9 million, which would bring the Free Library's total budget to $76.7 million. Confirming a precise number for the Free Library’s benefits costs requires more detailed math and accurate data, but one external source suggests it’s a reasonable guess. Every year, the Library Journal releases an index comparing public libraries across the country. Its most recent release in December 2020 compared library systems based on data from FY2018, and the Free Library lists staff costs of $40.7 million for salaries and $37 million for benefits, for a total staff cost of $77.7 million. So, while I can’t be sure of the exact personnel cost for the Free Library in FY20, it seems reasonable that total personnel costs could be somewhere in the ballpark of $75 million instead of the $40.6 million listed.
This guestimate shows how much current budget reporting deflates the actual cost of operating a department, especially the City’s two large public safety departments. If the same 76% benefit ratio is applied, the Police Department's personnel budget would be $1.3 billion instead of $774 million and the Fire Department's personnel budget would be $506.7 million instead of $331 million. Again, the actual math on these costs would be tricky, especially since Pew and the Economy League detailed more than a decade ago the cost of health insurance per person for Police and Fire was 34 percent and 48 percent higher, respectively, than the cost for other City workers, These types of substantial differences in costs between departments are even more reason why the budget should allocate costs by department.
But let’s assume that the budget showed all personnel costs by department, would we have a “full” picture of how much it costs to run a department? Not exactly. Even if a department budget included precise amounts for benefits, the costs for centralized functions like fleet management, human resources, or information technology are not allocated by department. Without including full cost allocation for these and other indirect operating costs, we simply don’t know the true costs to run a department or deliver a service.
After spending more than 1,300 words explaining how the City's budget does not paint a complete picture of what it costs to run City government, I must acknowledge the elephant in the room: does it even matter? If the budget showed full costs by department, would anyone besides a small group of policy wonks even care? As a friend, mentor, and local government expert put it to me, “the average Philadelphian is interested in how many police officers or librarians are in a department to provide public safety or deliver services and less about the healthcare and other compensation costs.” I agree completely with that sentiment, and that is why we need to know how much these departments cost to operate.
Our public discourse about City services should be rooted in a value proposition: do we believe the quality of services and the outcomes they produce are worth our investment of public resources? Considering the City’s budget has increased more than $1 billion in the last five years without tangibly moving the needle on Philadelphia’s biggest challenges, it is imperative that we clarify the levels of service we want and how much we are willing to spend to get there. In a city with an extensive list of long-term challenges, how can we decide on the most effective solutions if we can't compare costs, effectiveness, and make trade-offs? We can’t even begin to have an outcomes-focused public conversation to address our challenges if we don’t know how much it truly costs to deliver services.
Let’s look at two charts that demonstrate how clear and accurate cost allocation and reporting could help shape our priorities. The chart below from Pew’s report on the costs of City government shows how much we spend per capita on major city services. Based on the well-worn cliché which contends "a budget is a moral document" that reflects our values, we value Police 750% more than Parks and Recreation ($517 vs. $69 per capita), and 975% more than Libraries ($517 vs. $53 per capita).
Given this level of investment, are we satisfied with quality of services and the outcomes they produce? If not, are there investments we can make outside of the Police Department to make our neighborhoods safer? Portions of the $155 million for violence prevention in the FY2021 budget will support evidence-based strategies such as increasing green space, hiring outreach workers, or funding structured youth programming, resulting in more resources for Parks and Recreation and the Free Library. Hopefully, the newly formed Violence Prevention and Opportunity review committee evaluation of these investments can be an important step towards a budget more focused on outcomes than staffing. And if we knew how much it costs to run each City department, it could facilitate meaningful public discourse about where and how we invest our hard-earned public resources.
The second chart, based on data from 2015, shows the average wages for City employees who deliver most services. Police officers and firefighters earned more than $80,000 per year in wages, compared to less than $60,000 per year for most other City employees. Add in the average cost of benefits, and it is plausible that the total cost of one public safety employee could equal 1.5 or more non-public safety employee. If we had an idea of how much an average police officer or a librarian costs, for example, it would make for robust, outcomes-focused civic engagement for the budget.
Let’s say that the cost of 1 police officer or 1 firefighter is equivalent to 1.5 employees for all other City departments. In this scenario, a reduction in 50 police officers and a reduction of 50 firefighters could fund 75 new employees for the Free Library and 75 new employees for Parks and Recreation who could be force multipliers for the recent violence prevention investments. Using the same illustrative assumptions, 50 less police officers could fund 75 crisis workers who could provide the type of socio-medical response required for people dealing with “substance abuse, homelessness or other personal crises” that police officers are not trained to provide. Considering these types of incidents are almost one of every ten 911 calls from the public, this strategy could help reduce calls for service for the Police Department and free up police officers to spend more time preventing violent crime. Continuing with this scenario, 50 less firefighters could fund 75 more sanitation workers to help collect trash on time and ease the burden on the Streets Department’s workforce.
I am not advocating for these specific public policy strategies, but I believe that we – residents, elected officials, activists…everyone – should have clear, straightforward information about how much City government truly costs so we can thoughtfully and honestly make public investments in our present and our future. So, what should we do about it? There are, in fact, some simple steps we can take to foster more “truth in budgeting” in Philadelphia:
Departmental Cost Allocation for People: Allocate the cost for pensions, healthcare, and other fringe benefits by department in the City's budget and other financial reports. This would give the public a clear, simple, and transparent way to understand how $2 of every $3 for the operating budget is spent.
Departmental Cost Allocation for Centralized Services: The City should present departmental budgets with a full cost allocation model that combines direct costs like salaries or supplies with indirect or “overhead” costs like facility maintenance or information technology. Among many benefits, the most crucial outcome of cost allocation would be to holistically account for the cost of city services being provided.
Job Function Cost Overviews: Develop a simple method for displaying the average cost for core job functions of City government, such as police officers, firefighters, librarians, social workers, and sanitation workers. This simple one-pager details the full cost of a police officer in Chicago of $149,362, including benefits and indirect supervisory costs. These overviews would be the answer to “how much does a librarian cost?” and could be a valuable civic engagement tool for educating the public on the budget.
These changes alone won’t solve Philadelphia’s most challenging problems, but they can serve as vehicles on the road to finding solutions that are rooted in transparency, collaboration, and civic engagement. We have to start these conversations somewhere, and we might as well start with a budget that does the things it is supposed to do.
[1] Some departments have estimated employee benefits costs for specific programs, reported on form 71-53P. This information, however, is not a full illustration of department-wide employee benefits costs and many departments do not yet provide this type of program-level detail.