Troy runs Information Technology as an internal service fund, so 21 years of audited financial reports show its full cost as a single line of business — and the city's Adopted Budgets show what that money actually runs. Here is what the numbers say, FY2004 through FY2026.
Troy does not bury IT inside the general fund. It runs as an internal service fund (ISF): a self-contained business that bills every other city department for the technology services it provides. That structure is why 21 years of ACFRs each carry IT as its own set of financial statements — and why "what does Troy spend on IT?" has a clean answer.
The measure used here is the total operating expenses of the IT internal service fund — personnel, supplies, and other service charges, plus depreciation. It excludes capital-asset purchases, which flow through the fund's balance sheet rather than its operating statement (those are covered in Section 3).
Over the full span IT spending grew a little faster than inflation — the US CPI rose about 67% across FY2005–FY2025 against IT's 78% — but not dramatically so. The real change is not in the dollar total; it is in what those dollars now have to cover.
| Fiscal year | Operating expenses | YoY | Operating revenues |
|---|
Every Adopted Budget's IT section carries a Performance Indicators table: how many personal computers, printers, servers, and software applications the department keeps running. The tables overlap two-to-four years between budgets, so most years are cross-checked against several documents. Stitched together, 18 budgets cover FY2004–FY2026.
The same ~9–10-person team runs 2.4× as many applications as it did in 2004 — on a device fleet that has barely grown.
That is the core finding of the budget data. A city's IT department is not judged by how many PCs it owns — that number tracks headcount, and Troy's is stable. It is judged by the surface area it has to keep running: financial systems, permitting, assessing, GIS, the public website, email, telecom, and dozens of line-of-business applications. That surface area has more than doubled while staffing held flat.
| FY | Applications | PCs | Printers | Servers | Help-desk requests |
|---|
"IT asset spend" has two parts in the budgets, and they behave very differently. Capital — new servers, major software, infrastructure — is the adopted IT capital budget, and it is lumpy: it spikes when a big project lands and falls back otherwise. Equipment maintenance is the running cost of keeping the existing fleet alive, and it is comparatively steady.
The accounting here is genuinely subtle, and worth stating plainly: the adopted capital budget is the consistent, comparable series — but in most years the IT fund's own statements show $0 of capital actually booked, because historically the city funded IT capital through operating transfers from its Capital Projects Fund rather than through the ISF's capital line. Only five years (FY2007, FY2008, FY2017, FY2023, FY2024) show non-zero ISF capital actuals. The bars below are therefore "what the city planned to invest," and should be read as such.
isf_capital_actual column of it_assets.csv.
| FY | IT capital budget (adopted) | ISF capital actually booked | Equipment maintenance |
|---|
This is a working analytical record, not an audit. Everything is reproducible from the cot-it repository — the source PDFs, the extraction scripts, and the resulting CSVs are all there.
Two pdfplumber-based extractors do the work. extract_it_spend.py finds the Internal Service Funds combining statement in each ACFR and pulls the IT column's operating expenses and revenues. extract_it_assets.py encodes every Performance-Indicators cell read from the budgets — each tagged with its source budget and column type — then resolves each fiscal year to a single best value (Actual beats Projected beats Budget; the newest budget wins ties) and prints the full provenance.