A Chicago resident opens a crime dashboard and sees a sharp-looking change in reported incidents. What does that change actually measure? For a useful Chicago neighborhood safety comparison, begin with one community area, one offense category and one defined time window. Our archived 2018 campaign materials raised questions about community safety. Here we approach the subject through public records, without turning a neighborhood into a label or treating reported cases as a complete measure of daily life.
Question one: what exactly was counted?
A report, an offense, an arrest and a clearance are different units. The Chicago Office of Inspector General's reported-crime dashboard lets readers examine crime counts, categories, locations and clearance measures. It also distinguishes the number of police case reports from the number of offenses. A single case report can contain more than one offense.
Imagine a completely hypothetical month in which a community area has 110 reported offenses and 100 case reports. The difference is possible because some reports involve multiple offenses. Neither number would tell us how many residents were arrested. Nor would it show every incident residents experienced but never reported.
The dashboard has a separate convention for homicides, counted by victims rather than the same case-report logic. Its clearance-rate measure is also specific: cases cleared during a period can relate to offenses from earlier dates. Interpreting a clearance percentage as the share of that month's new cases immediately solved would distort the measure.
Before using any headline total, answer these questions. They help establish what the chart can support:
- Is the number a count of reports, individual offenses, arrests or clearances?
- Which offense category and classification system have been selected?
- Does the source combine categories or display them separately?
- Is the comparison showing raw numbers, a percentage change or a population-adjusted rate?
There is no universal safety score hiding inside one of those fields. A change in recorded theft reports, for example, does not automatically describe changes in other offenses or residents' perceptions.
Question two: are the place and time genuinely comparable?
Chicago uses community areas, police districts and police beats for different administrative purposes. A police beat is a patrol geography; a community area is a different boundary system. If one report uses beats and another uses community areas, lining up two totals without checking boundaries can produce a false contrast.
Time windows need similar discipline. Comparing twelve complete months with the first nine months of a different year is misleading, even when both figures look precise. Seasonal patterns, changing data entry and differences in reporting can affect short-term readings.
Use this dashboard reading card before drafting a sentence about a neighborhood. It turns an eye-catching chart into a verifiable statement:
| Question | Field to inspect | Common mistake | More careful wording |
|---|---|---|---|
| Which incidents? | Offense type and reports/offenses setting | Treating all categories as interchangeable | “Reported theft offenses” |
| Where? | Community area, police district or beat | Comparing different boundary systems | “Within the selected community area” |
| When? | Start date, end date, update date | Comparing unequal windows | “During the same calendar months” |
| How measured? | Count, change or clearance rate | Calling a raw count a safety ranking | “The number of recorded reports changed” |
Large and small populations also complicate comparisons. A greater number of reports can partly reflect a larger number of residents or visitors. Even a per-resident rate does not account for every difference in daytime activity, reporting practices or risk.
Question three: what remains outside the dataset?
The City's published crime records and mapping data describe reported incidents, subject to revisions. The underlying public dataset excludes the most recent seven days, uses approximate block locations to protect victims' privacy and warns that preliminary classifications may change. A map point should never be treated as an exact incident address.
The OIG dashboard, introduced in January 2026, draws from police databases beginning in 2013 and typically refreshes monthly. It displays the last update date so readers can tell how recent its figures are. Those dates matter when a news report and a public dashboard show slightly different totals.
An incident that was never reported cannot appear in a report-based count. Conversely, one recorded event does not define everyone living nearby. Personal safety involves context and experiences that this dataset was never intended to measure.
A short comparison lab
Try one narrow question, such as whether reported vehicle thefts changed in a selected community area between the same months of two years. This is a method exercise, not a claim about any particular Chicago location. Record your settings before you interpret a chart.
The following routine keeps the comparison reproducible. Complete each check before sharing the conclusion:
- Identify the data owner and read the definition of the measure you selected.
- Save the last-update date and choose equal start and end periods.
- Keep the same community area and offense classification in both views.
- Check whether the display is a count, percent change or clearance rate.
- Describe only the change shown, while noting missing or provisional information.
Then ask whether a reader could rebuild the comparison from your settings. If you cannot identify the filter choices, the finding needs more work. If categories or boundaries changed between extracts, explain that limitation rather than supplying a dramatic number.
Speak about places without ranking their people
Public safety data can guide specific questions for local meetings, reporting and research. They cannot promise that one street is safe or unsafe for an individual. Our historical interest in Chicago communities gives us a reason to encourage careful use of these records, not to label residents or neighborhoods.
When you share a comparison, include the place definition, offense, dates and source-update date in the same paragraph. Readers deserve enough context to distinguish a documented change in reported incidents from a broad judgment about community life.