Brian Keegan is a regular opinion columnist for Boulder Reporting Lab. His “Charting Boulder” column uses public data to make sense of how the city is changing — from housing and politics to income and population — with clear explanations and a focus on equity.
Boulder’s budget problems are not going away, but maybe you have the solution. Its approved 2026 budget was $521 million. The city manager’s recommended 2027 budget totals $552.6 million and closes a $6.3 million gap in the General Fund by eliminating 24 positions, trimming services, raising fees and moving some costs onto voter-dedicated funds.
The headline causes of Boulder’s budget woes have been flat sales tax revenue and declining property tax revenue following changes in state law. The city council will vote on the 2027 budget at its Oct. 15 meeting. One of the biggest challenges councilmembers face is that much of the city’s money cannot simply be shifted around. Nearly two-thirds of the 2027 budget (64%) is restricted or otherwise committed.
Open space, utilities, grants and capital projects already have claims on much of that money before staff or council make budget decisions. Funding that voters have earmarked for open space, parks and recreation, and the arts generally cannot be moved to plug holes elsewhere in the budget.
However, expenses that serve those voter-approved purposes can be shifted from the General Fund to those dedicated funds. The recommended 2027 budget, for example, shifts about $570,000 of General Fund costs to funds supported by the open space and sugary drink taxes. That includes about $180,000 for the wildland fire crew that would move from the fire department’s budget into the open space budget.
The General Fund is what most people imagine when they think of a city budget. It is the discretionary slice of the budget where police, fire, human services and much of city government compete for the same pool of money raised primarily from sales taxes but also from property taxes, fees and fines. In 2026, the General Fund totaled about $194 million and faced a $7.5 million gap. The council closed that gap in October 2025 with a mix of cuts and new fees. The recommended 2027 General Fund is about $200.5 million.
Inspired by a 2010 New York Times interactive for balancing the federal budget, I used Claude Opus 5.5 to vibe code an analogous budget-balancing tool for Boulder’s 2027 budget. It is a coarse simulation of the biggest budget levers and omits minor items, one-time charges and complex accounting details best left to professionals.
Blue ticks mark the choices in the city manager’s recommended budget, allowing you to compare your choices with the proposal before council. A provenance file traces the source of every number.
Your mission: Close the estimated $6.3 million gap in the 2027 General Fund. (Click into the interactive graphic below and scroll through your options to build your budget.)
The sliders start at a baseline that removes the city manager’s proposed fixes, leaving the $6.3 million gap open. Follow the blue ticks and you’ll close it using the city manager’s recommendations.
Or make different choices.
Like council, you will not be able to submit your budget until the deficit is gone. Once you close the gap through cuts and revenue changes, we’ll ask you a little about who you are.
The results are not meant to be a statistically valid survey. Neither were the city’s “Fund Our Future” conversations shared with council in June. But the responses could offer a window into the trade-offs Boulder residents are willing to make.
Behind the budget experiment
I would be remiss as your friendly neighborhood information scientist if I did not remind you that all data have politics: This widget is not neutral and encodes the assumptions and biases of its human and AI creators.
What do I mean?
A slider set to 0% does not simply carry the 2026 values into 2027. Instead, the tool’s starting point is calculated by reversing the proposed 2027 changes and assigning them to the sliders’ categories to produce the opening $6.3 million deficit.
Claude persuaded me to use this complicated “reverse engineering” budget model to establish the 2027 baseline. If you want to revisit that decision, pull requests — proposed code changes — are welcome on the column’s GitHub repository.
Other faulty assumptions were wholly my own. An earlier draft of this tool only allowed the revenue sliders to increase — taxes and fees could only go up. But budgets can be balanced with lower revenues and expenditures too. The tool now lets users decrease or shift revenue sources and then evaluate the trade-offs. (Shout out to all the other SimCity fans out there.)
The data you submit are anonymized (no user-agent, IP address or cookie data are stored with your responses), and the tool includes built-in logic to filter out “enthusiastic” responders who submit multiple responses in a short period of time. Comparing readers’ budget choices with their demographics may reveal interesting patterns, which I’ll analyze in a subsequent column.
One of the slipperiest moves common across science and policymaking alike is building a story around your favorite finding after you have seen the data. One cure for this kind of cherry-picking is to write your questions and analysis strategy down beforehand, a practice scientists call “preregistration.”
For example, I hypothesize that homeowners and older residents will be overrepresented among respondents compared with Boulder’s general population. I also hypothesize that these respondents will be more likely to propose budgets that cut expenses rather than raise new revenue.
I’ve started to outline the analyses I will run so that my analysis of the submissions does not highlight only the most convenient findings. Before a single response has been submitted, I have committed to exploring these preregistered questions in a follow-up column reporting the results:
- Cutting expenditures or raising revenue. How does the split between spending cuts and revenue increases track demographics like housing, tenure, age and income?
- Drastic or incremental budget reforms. How many people try to max out changes in revenue and expenditures rather than make incremental changes?
- Consensus versus council. Which General Fund spending and revenue levers draw the most support, and how do those choices compare with the choices in the city manager’s recommended budget?
- Extrapolating staffing impacts. How would staffing levels in the city change, particularly under the strongest cost-cutting scenarios?
I welcome your own questions and hypotheses in the comments below!
The code for replicating these analyses can be found on the Charting Boulder GitHub.
It’s not lost on me that most people have cut the police budget. The value for what we get just isn’t there, perhaps they should provide a better product.
Very nice exercise and I appreciate the amount of effort you put into this. Cut planning staff and communications department, city attorney, increase police and fire budget, increase rec center and non-professional city staff.
Fire, police, other public safety services should not be cut. Unfortunately need to reduce other areas and increase marijuana tax.
Hey Brian, can you help balance the budget for SBC flood mitigation project? Seems like we’re about $75m over original estimates and no one from the city, WRAB, the progressives care to present the ledger… do you know what they’re hiding?
Without increasing any revenue stream my submission balances the budget by focusing on superfluous spending. Hardest hit would be the City Managers Office with all the boutique programs they’ve ramped up over the last few years.