Part 3 of 7 in The Overtime Machine

The Lede

Every year, Philadelphia's budget books say the same thing: “Overtime — $X million.” They do not say — because no budget document can say — who inside each department actually receives those dollars. They do not show that in Police, the OT distribution is concentrated at the top. They do not show that in Fire, the same pattern repeats. They do not show that 42 cents of every overtime dollar go to roughly one in ten of the workers who earned any overtime at all.

The curve above fills part of that gap.

The Inequality Nobody Talks About

Here is a number worth sitting with: 0.619. That is the Gini coefficient for overtime distribution at the City of Philadelphia in 2025. A Gini coefficient measures inequality — zero means everyone gets the same, one means one person gets everything. The United States household income inequality Gini is about 0.49, per the Census Bureau's most recent release. Philadelphia's overtime inequality is higher than that — meaning how dollars flow inside one employer is more unequal than how dollars flow across the entire US economy. I did not expect that comparison when I started pulling the data.

Now here is a second number worth sitting with: 41.7%. That is the share of all overtime dollars that went to the top 10% of overtime earners in 2025. Not 10% of workers — 10% of overtime earners. The city has about 33,600 salaried employees on its payroll in 2025. Of those, roughly 22,000 earned any overtime at all in the year. The top 10% of overtime earners — roughly 2,200 people — captured 42 cents of every dollar the city paid in overtime last year.

The Same Pattern Inside Every Department

One way to understand this problem is to look at it department by department. The chart below shows the overtime concentration within four representative departments: Police, Fire, Water, and Streets. Each panel shows the same Lorenz curve analysis applied to a single department's overtime distribution. The Police department in 2025 had a Gini of 0.55 for overtime. Fire was at 0.47. Water was at 0.57. Streets was at 0.54. Every department does this.

How the Machine Works

The reason, I think, is structural: overtime opportunities in city government are not distributed by managers scanning a list. They appear to be distributed through a system built on accumulated seniority — the workers who have been there longest get first pick of the shifts that generate the most overtime, and new workers get what is left. I do not have the shift-allocation data to prove this. But the pattern is what tenure-based scheduling would produce.

The chart below makes the pattern visible. Each dot is a department in a given year, positioned by its average base pay and average total compensation. Departments above the diagonal are paying their workers more in total than the base salary explains — the gap is overtime, longevity, and other supplemental pay. Police and Fire pull furthest from the line. They are not just high-overtime departments. They are departments where a small number of senior workers have structured their financial lives around shift swaps and double-time opportunities.

The Two-Tier Workforce

What the overtime data reveals is that Philadelphia has effectively two workforces wearing the same uniform. The first tier is senior workers who have been in their departments long enough to hold the shifts that generate the most overtime. They have structured their lives around these shifts. Their total compensation reflects it — overtime can add 50%, 100%, or more to their effective pay. They are not working more hours because the work requires it. They are working specific hours because those are the hours that pay.

The second tier is everyone else — newer workers who cannot access those shifts, who earn primarily base salary, and who are effectively subsidizing the overtime system through their lower compensation. This is not a complaint about new workers being less productive. In many cases they are more productive — more physically capable, more current in their training, more adaptable. They simply cannot get the shifts that pay.

The gap between these two groups is not a performance difference. It is a seniority difference that has been converted into a compensation difference through the structure of how overtime is allocated.

The right tail of the chart is where overtime stops being a supplement to base pay and starts being the dominant half of compensation. The tail is small. That is what a system looks like when it rewards seniority without limit.

The Budget Question

Budget documents treat overtime as a line item. They show total overtime spending by department and call it a cost management problem. What they do not show is the distribution — the fact that overtime is not a general supplement to city worker pay, it is a parallel payroll flowing to a small, senior group within each department.

When someone says the city needs to hire more workers, what they are often proposing is an addition to the second tier — new workers who will do the base work at base pay while the first tier continues to work the overtime-eligible shifts. The total payroll grows. The overtime concentration continues. The problem is not being solved.

Some cities have tried to fix this. Rotating overtime lists that give newer workers a chance at high-overtime shifts. Individual caps on overtime earnings that prevent any single worker from capturing too large a share. These are structural interventions, not cost-cutting exercises — they change who gets the overtime, not how much overtime gets worked.

The Gini coefficient for citywide overtime has grown from 0.585 in 2023 to 0.619 in 2025. That is not a large move in percentage terms, but it is moving in one direction. The concentration is increasing. The gap between those at the top and everyone else in the overtime distribution is getting wider.

The top decile takes 42 cents of every overtime dollar. The Gini has gone from 0.585 to 0.619 in two years. The budget line will look the same next year. The concentration behind it will not.

Notes, Sources, and Methodology

Source: City of Philadelphia Employee Earnings dataset, published via OpenDataPhilly. The dataset covers every employee on the city payroll, with quarterly earnings broken out by category — base, overtime, longevity, post-separation, and miscellaneous.

Scope: Salaried employees, calendar years 2023–2025. Overtime figures are annualized from the four quarterly records per employee, then ranked to produce the Lorenz curves and Gini coefficients above. The citywide headcount of roughly 33,600 counts unique salaried employees who received any pay during 2025; the 22,000 OT earner figure counts unique employees whose annualized overtime was positive in 2025. Part-time, seasonal, and hourly workers are excluded because their quarterly records do not represent a full year of full-time pay.

Caveat: The data reflects earnings, not budgeted headcount. Employees on extended unpaid leave are excluded if they received zero pay in all four quarters of the year. Department-level Gini figures can be noisy in small departments; figures shown are for departments with a stable salaried headcount large enough to support the calculation. The US household-income Gini figure is from the US Census Bureau, “Income in the United States: 2024” (P60-286, published 2025-09-09).