The Case for FP&A Analysts in City Government

Every city department has people who do real work. The garbage men who collect the trash. The judicial clerks who file the motions. The license approvers who stamp the permits. These are the people who make the city function, whose hands touch the services that residents depend on, whose labor is visible and tangible and necessary.

So here's my proposal, and I want you to not immediately dismiss it: each Philadelphia department should hire people to do "fake work" — analysts who sit at computers, crunch numbers, and ask questions like "why is this vendor charging 30% more than market rate?" and "why does overtime spike every August like clockwork?" Sounds pointless, right? Sounds like the kind of people who schedule meetings about scheduling meetings. But I've spent three years on the outside of city government with Python and publicly available data, and I'm about to convince you — I think — that this "fake work" might be the most important work a department does.

It is a well-known aphorism that "knowledge is power." In today's world, knowledge is code for data. But data is just numbers on a page — until someone with the skills and the wisdom to massage those numbers can make them dance in a pattern that reveals insights which were previously hidden. That's what FP&A analysts do. Call them what you will — data scientists, number crunchers, quantitative oracles — they're the people who turn data into knowledge.

Let me try to explain why I think these analysts are the most missing, high-value function in Philadelphia city government.

The Arithmetic

Philadelphia's operating budget is roughly $7 billion. That's a number that gets said often enough to lose its weight, so let me put it another way: it's roughly $4,500 for every resident of the city, every year, in perpetuity. Now here's a proposal: take twenty of the largest departments — the ones with the biggest budgets and the most employees — and put one financial analyst in each one. Not a consultant. Not an auditor. An actual employee, someone embedded in the department, whose job is to watch the money the way a GPS watches the road.

The total compensation for someone like this — salary, benefits, the overhead — runs somewhere between $100,000 and $150,000 per year. That's a senior analyst or a quantitatively-minded MPP graduate, someone who can actually read a dataset and form an opinion. Twenty departments times one analyst times $125,000 on average: we're talking about $2.5 million annually. As a share of the city budget, it's less than half a percent. A rounding error. A rounding error that might be the best money the city never spends.

But here's the part that makes me actually laugh when I first ran the math: the analysts only need to find 0.04% of the budget in efficiency improvements to justify themselves. Not 1%. Not 5%. Point-zero-four percent. That's the bar. That's the entire ROI case. Find that much, and you've paid for yourselves thirty times over. And I've — from the outside, with limited data, with no badge, with no database login — I've found patterns suggesting that 1-2% improvements are realistic. If someone inside the building found half of what I've found, the city saves roughly $35 million on a $7 billion base. The analysts would pay for themselves thirty times over.

The analysts don't replace the workers who do the real work. They make them more effective. They find the places where the system is leaking money, where the processes are slow, where the spend is misaligned with the mission. They don't fire the garbage men — they figure out why the routes are inefficient, why the overtime is spiking, why this vendor charges 30% more than market rate for the same service. They make the people doing the real work faster, cheaper, and better supported.

The Gap

The Mayor's budget office and City Council allocate the budget before the fiscal year starts — that's a plan, a set of numbers on a spreadsheet. The Controller's office audits what happened after the fiscal year ends — that's history, written in past tense. Between those two functions sits the entire fiscal year, during which decisions get made, invoices get paid, contracts get signed, overtime gets authorized, and nobody — nobody — is watching. Nobody asks whether the department is tracking to plan. Nobody asks why vendor X charges 30% more than vendor Y for the exact same service. Nobody asks why overtime is doing something different this August than it did last August. Nobody asks anything until after the money's gone.

This is the gap. And it matters.

FP&A analysts fill exactly this gap. They are the people whose job is to watch in real time — to track budget vs. actuals as the fiscal year unfolds, to flag problems before they become year-end surprises, to find the places where money is being spent badly or inefficiently or in ways that don't serve the mission. They are the GPS. Not the rearview mirror. The federal government figured this out decades ago. Every Fortune 500 company has an FP&A function. It's not exotic. It's not ideological. It's just what you do if you're serious about managing money.

And here's the distinction that matters: FP&A is not auditing. The Controller's office audits — they check the rearview mirror, they review what happened, they issue reports after the fiscal year closes. FP&A is real-time. It's the difference between a rearview mirror and a GPS. You need both. But you can't steer a car with a rearview mirror, no matter how good the mirror is. And the auditor's report, however thorough, doesn't help you fix the problem while the fiscal year is still running.

What I've Found

I want to be clear about what I know and what I don't. I've spent three years writing about Philadelphia's city budget — overtime patterns, department-level spending, payroll growth, vendor contracts, staffing efficiency. I've built datasets from publicly available sources, FOIA requests, and open data portals. I've found patterns that suggest meaningful inefficiency: overtime spikes that follow predictable seasonal patterns in some departments but not others, payroll growth that outpaces inflation in specific areas, vendor pricing that varies dramatically across similar services, permit approval times that range from days to months for functionally identical requests.

I've written about some of this. The overtime addiction essay. The department scatter data. The payroll growth analysis. The patterns are there, and they suggest that a person with real access — not me on the outside with limited data — could find more. Much more.

One person, some Python code, publicly available data, limited access to the building. And I've found patterns suggesting 1% to 2% inefficiency concentrations in the data I can access. Imagine what someone inside with a badge and a database login could find. Imagine what they could fix.

The Objections

I can hear the objections forming. Let me address them.

The first objection is that this is just more bureaucracy. FP&A analysts don't do real work — they run spreadsheets and write reports. But this misunderstands what the function actually is. FP&A analysts aren't a new layer of administration. They're the people who find the problems that the auditors miss and the patterns that the budget allocators don't see until after the decisions are made. Every large private organization has this function, not because they enjoy bureaucracy, but because they require it to stay functional. It's not bureaucracy. It's operational intelligence.

The second objection is that the Controller's office already audits. Why do we need both? But this is exactly the point. The Controller's office audits ex post — they check what happened after the fiscal year ends. FP&A is real-time, continuous, during the fiscal year while there's still time to adjust. The Controller's audit is the rearview mirror. FP&A is the GPS. You need both, but they serve different functions, and you can't use one in place of the other. A good GPS doesn't replace the rearview mirror — it tells you where you're going. The audit tells you where you've been.

The third objection is that the city can't afford this. But $2-3 million annually across twenty departments is 0.04% of a $7 billion budget. The analysts only need to find 0.04% efficiency improvement to break even — to pay for themselves. The question isn't "can we afford to spend $2.5 million on financial analysts?" The question is "can we afford to continue operating without any real-time financial visibility into a $7 billion budget?" One of those questions has an obvious answer. The other one we just keep not asking.

The fourth objection — the one that matters most — is that this will just be used as a pretext to cut services and fire workers. That FP&A analysts are just the advance team for austerity, finding the "efficiencies" that really mean fewer services and layoffs. This is the right objection to have. And it's wrong for this reason: efficiency doesn't have to mean cutting. Sometimes the right financial answer is to spend more — to invest in better tools for the workers doing the real work, to increase training and compensation for the staff whose labor is the actual service, to fund the programs that are actually working and cut the ones that aren't. The Laffer curve logic applies here: sometimes the most efficient move is to raise fees or taxes where the math supports it, not to cut spending. FP&A analysts find the truth, whatever it is. That's the point. The point isn't to find cuts. The point is to know what's actually happening so you can make the actual decision that the actual situation actually requires.

Turn on the Lights

The efficiency gains don't have to mean shrinking government. They can mean making government work — more effectively, more intelligently, with better information flowing to the people making decisions. The goal isn't to make government smaller. The goal is to make it work. And you can't make something work if you don't know how it's performing.

The cost of this is roughly $2.5 million annually across twenty departments — a rounding error on a $7 billion budget. The return is the possibility of finding out whether the $7 billion is actually being spent well. The federal government figured this out decades ago. Every Fortune 500 company has an FP&A team. Mid-market companies figured it out too. City governments are, for reasons that escape me, still flying blind.

It's not a partisan position. It's not an ideological argument. It's just accounting. The hardest part of this work isn’t the math. It’s accepting that the people inside the institution know things that people outside don’t. I learned that the hard way, in a project retrospective where a more experienced colleague corrected my framing—the reflection is my piece on the developer who stopped me mid-sentence. Both lessons point the same way: institutional knowledge is the missing function.