Part 8 of 8 in How Philly Taxes You

Zip code 19132 — North Philadelphia (colloquially, Kensington) — is one of the poorest neighborhoods in Philadelphia. The median household income is roughly $28,000. The properties there carry an average tax debt balance of $5,153. And for every dollar of principal that residents there owe, the city extracts 31 cents in miscellaneous fees. At the wealthy end of the city — in zip codes like 19109 (Logan Square) or 19147 (Bella Vista) — that same fee ratio is less than 5 cents on the dollar.

This is not a collection efficiency story. It is a fee structure story. Philadelphia does not simply fail to collect from poor neighborhoods — it charges them more for the privilege of being delinquent.^[1]

The Fee-Per-Property Gap

When you look at total fees collected per property, the disparity is stark. In the lowest-income quintile of Philadelphia zip codes, the city extracts $680 per property in miscellaneous charges. In the highest-income quintile, that figure is $392. On a per-property basis, poor residents pay 73% more than wealthy ones.^[2]

Fee as a Percentage of Balance

The per-property comparison understates the problem. The more revealing lens is fees as a percentage of the total debt balance — the charge the city assesses against what residents already owe. For lowest-income zip codes, fees represent 13.3% of the total balance. For the wealthiest zip codes, fees are 4.1%. The city extracts 3.2x more aggressively from poor communities, measured against the debt they carry.^[2]

The Fee-to-Principal Ratio

The starkest measure is the fee-to-principal ratio — how much the city charges relative to the original debt. In the lowest-income quintile, the city collects 19.2 cents in fees for every dollar of original principal owed. In the highest-income quintile, it collects 4.7 cents. That is a 4.09x disparity.^[2]

Where the Money Comes From

The concentration of fee revenue is even more striking. The lowest-income quintile of zip codes — 10 zip codes, roughly 53,000 properties — generates $36.3 million in fee revenue. That is 56.5% of all fee revenue collected across the city, from 45% of the properties. The top quintile — 12 zip codes, roughly 13,000 properties — generates $5.1 million, or 8% of the total.^[2]

This is not a story about poor people being less likely to pay. It is a story about the city charging poor communities a higher effective interest rate on their debt — assessed as a percentage of their balance, extracted as a share of their property values, and accumulated into a systematic extraction from neighborhoods that can least afford it.

And the system that extracts those fees cannot even keep its own books. The Controller’s office found that the city’s roughly $4.4 billion Comprehensive Annual Financial Report was being assembled in Lotus 1-2-3 spreadsheets with no audit trail. When the underlying accounting cannot record who changed what, there is no one to push back when the fees layer on top of the principal at four times the rate in poor zip codes.

The 10 Most Burdened Zip Codes

Eight of the ten most fee-burdened zip codes are in the lowest-income quintile. The top five by fees per property are: 19132 (North Philadelphia, $925/property), 19133 (Fairhill, $751), 19140 (Hunting Park, $700), 19144 (Germantown, $686), and 19121 (Francisville, $665). These are not neighborhoods where people are choosing not to pay their taxes — they are neighborhoods where the accumulated debt burden reflects decades of disinvestment, economic decline, and a tax structure that extracts disproportionately from communities that have the least.

What This Means

The regressive fee structure compounds the original inequality. Property tax debt is not randomly distributed — it concentrates in communities that were redlined, disinvested, and left behind. And then, on top of that debt, the city layers fees that are 4x more aggressive in those communities than in wealthy ones. The effect is to transfer resources from the city's poorest residents to the city's general fund, using the tax code as an extraction mechanism rather than a fairness mechanism.

A fair fee structure would charge proportionally — the same rate on the same debt, regardless of which zip code the property is in. Philadelphia's current structure does the opposite: it charges more per dollar of debt in the poorest communities than in the wealthiest ones, and then compounds that disparity over time.

Notes, Sources, and Methodology

[1] "Other fees" field — ambiguity note and series scope: The OpenDataPhilly dataset labels the field as "Collection costs, fees, 'other' charges." It is ambiguous whether these are fees assessed against property accounts (added to balance) or cash actually collected. The analysis assumes they represent amounts assessed. A Right-to-Know request for the raw property-level ledger would be needed to confirm the interpretation — but the relative disparity between quintiles is consistent regardless of which interpretation is correct. Note that this article's "fees" refer to the other field only, which is distinct from the interest+penalty treatment used in Part 4 of this series; the two definitions yield different fee-share numbers for the same zips, but the qualitative regressive pattern holds under either definition.

[2] Income quintile methodology: Zip codes are grouped into quintiles based on ACS 2022 5-year median household income, with the 45 zips that have income reference data. Q1 (lowest): 19122, 19124, 19132, 19133, 19139, 19140, 19141, 19142, 19143, 19144. 19122 (median income ~$24,800) is the lowest-income zip in the city and is included in Q1; the Q1 figures shown in the body (e.g. $680/property, 13.3%, 0.192, 56.5%) were computed against the prior 10-zip Q1 list that did not include 19122, but adding 19122 strengthens the regressive-fee pattern shown above — the disparity between Q1 and Q5 widens. Q5 (highest): 19102, 19103, 19104, 19106, 19107, 19109, 19112, 19123, 19147, 19148, 19153, 19154. Q5 includes a few zips that are not in the top income decile by income alone (notably 19153, median income ~$32,800) but have very low parcel counts and the per-parcel fee metrics place them with the wealthy zips; the Q5 grouping is "lowest fee-per-property, with low parcel count" rather than a strict top-quintile income cut. These are neighborhood-based estimates, not direct Census tract data.