Seasonal Variation in Police Incident Categories in San Francisco
1 Background
This project examines seasonal variation in reported police incidents in San Francisco during 2025 using the San Francisco Police Department Incident Reports dataset accessed through the DataSF AP]. Previous research suggests that crime does not occur uniformly throughout the year. Rather, seasonal changes in routine activities influence opportunities for crime and may contribute to temporal differences in the types of incidents reported (Cohen and Felson 1979; Andresen and Malleson 2013).
The purpose of this preliminary analysis is to explore whether police incident categories exhibit observable monthly variation prior to formal hypothesis testing. The final research project will evaluate whether the distribution of incident categories differs significantly across months using a chi-square test of independence.
1.1 Hypothesis
H1: The distribution of incident categories will differ across months because seasonal changes influence routine activities, public interactions, and opportunities for crime. Further, based on the existing literature, I expect incident categories to show statistically significant seasonal variation, with some categories occuring more frequently during particular months than others.
2 Initial Findings
2.1 Figure 1: Monthly Incident Reports
Figure 1 displays the total number of police incident reports recorded each month during 2025. Reported incidents fluctuate across the year rather than remaining constant, suggesting that temporal factors may influence reporting patterns. Although the figure does not identify which types of incidents contribute to these differences, it provides initial evidence that incident activity varies throughout the year.
2.2 Figure 2: Distribution of Incident Categories by Month
Figure 2 shows the monthly share of reports represented by each of the ten most frequently reported incident categories. Larceny theft accounts for the largest share in every month, although its relative frequency varies moderately across the year. Most other categories exhibit smaller month-to-month differences, with drug offenses showing a comparatively visible increase during the later months of 2025. The composition of reported incidents appears stable, but the observed variation supports further testing through a chi-square test of independence.
2.3 Table 1
| Incident Category | Number of Reports |
|---|---|
| Larceny Theft | 19135 |
| Assault | 6499 |
| Drug Offense | 6077 |
| Other Miscellaneous | 5720 |
| Malicious Mischief | 5698 |
| Warrant | 4563 |
| Motor Vehicle Theft | 3752 |
| Burglary | 3718 |
| Lost Property | 3399 |
| Fraud | 3241 |
Table 1 summarizes the frequency of reported incident categories in the dataset. Theft-related offenses account for a substantial proportion of all reported incidents, while several other categories occur less frequently. Understanding the overall distribution of incident categories provides context for subsequent analyses examining whether these frequencies change systematically across months.
3 Preliminary Conclusion
The preliminary analyses indicate that police incident reports exhibit observable temporal variation throughout the year. Monthly incident totals fluctuate modestly, and the proportional distribution of incident categories appears to change across months. Although these descriptive analyses cannot establish statistical significance, they suggest that seasonal variation may influence the composition of reported police incidents. The next phase of the project will formally evaluate this relationship using a chi-square test of independence.
4 References
Allaire, J. J., Charles Teague, Carlos Scheidegger, Yihui Xie, and Christophe Dervieux. 2022. “Quarto: R Interface to ’Quarto’ Markdown Publishing System.” https://CRAN.R-project.org/package=quarto.
Andresen, Martin A., and Nicolas Malleson. 2013. “Crime Seasonality and Its Variations Across Space.” Applied Geography 43: 25–35. https://doi.org/10.1016/j.apgeog.2013.06.007.
Cohen, Lawrence E., and Marcus Felson. 1979. “Social Change and Crime Rate Trends: A Routine Activity Approach.” American Sociological Review 44 (4): 588–608. https://www.jstor.org/stable/2094589.
R Core Team. 2022. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/.