An Independent Look at Albany Crime Data: Part 1

---

The FOIA crime data that we received had 78,094 incidents with dates between Jan. 1st, 2020, and Mar. 8th, 2025. In each case, incidents were tagged with a date, location type (e.g., Amusement Center, Parking Lot), crime classification (which is not regularly filled in), uniform crime report, and hundred block (general location, i.e., 700 Block of Park Ave, Madison Av & Lark St). A snippet of that data is provided below in Table 1. 

We define the entire list as "incidents" because under the heading UNIFORM CRIME REPORT, there are several incidents tagged as NO CRIME or ALL OTHER.  

Figure 2 (above) - UNIFORM CRIME REPORT between Jan. 1st, 2020, through Mar. 8th, 2025 after NO CRIME and ALL OTHER were removed. This data represents reported crimes. Percentages were renormalized, and the MISC. field contains the sum of all reports with less than 1%, i.e., MURDER.

The top 10 crimes and their counts are presented below in Table 2, with LARCENY (33.8%), CRIMINAL MISCHIEF (21.6%), and SIMPLE ASSAULT (12.0%) as the top 3 crimes.  

The first crime trend we examined was the total number of monthly crimes. We wanted to examine whether the crime rate had increased, decreased, or remained constant since January 2020. 

Figure 3 - All reported crimes after removing NO CRIME AND ALL OTHER. An upward trend, determined by a least squares regression fit, is observed along with cyclical highs near the middle of the year (June-August) and lows towards the end (November-February). 

Each point represents the count of all crimes reported in a given month. Reports from March 2025 were excluded because our records ended on March 8th.  We believe that the best way to visualize the overarching trend over 5 years of data is to use a linear regression trend line.  The value of this trend line is that it is showing you the overall trend and will not show dramatic changes if there are sudden short drops or increases.  Our trendline shows the changes in the data that are both significant and persistent over the 5-year time period we analyzed.     

Figure 3 shows a general rise in the overall number of reported crimes, with the most reported crimes occurring in mid-2022. Despite this spike, 2022 and 2023 had approximately the same number of reported crimes. We provide the total number of crimes reported for each calendar year here, excluding 2025, along with the percent change below. 

Table 3 (above):  Reported crimes by year and year-over-year change (%)

There was a significant drop in the number of reported crimes in 2024 relative to 2022 and 2023, which had approximately the same number of crimes. Our level of crime has returned to 2021 levels which, of course, was a COVID-19 pandemic year.  There is not enough data or information to suggest whether this trend will continue through 2025. Furthermore, this drop may be attributed to certain types of crimes, which we revisit later.

The cyclical data presented in Figure 3 also inspired us to group the crimes by month. To normalize for the large year-to-year fluctuations, the yearly percentage of crimes occurring in any given month was calculated rather than the average number of crimes. This maintained the seasonality aspect we were interested in exploring. Several factors may contribute to this change, such as city schools and college students living in Albany, warmer weather, and longer daylight. Winter months in Albany are colder, and travel may be more difficult in winter months, especially after a snow or ice storm. All of these factors could impact the rate at which crimes occur at outside locations. 

This led us to our first hypothesis: more crimes were occurring outside during warmer months. To test this assumption we grouped crime data based on LOCATION TYPE by splitting all crimes into a few broad groups: OUTSIDE (STREET, PARKING LOT, YARD, OTHER OUTSIDE LOCATION, GAS STATION, AUTO SALES LOT, FIELD WOODS, CONSTRUCTION SITE, PARK PLAYGROUND, LAKE WATERWAY, CEMETERY), INSIDE, and UNKNOWN; however, neither INSIDE or UNKNOWN were not analyzed as part of this article, but we may revisit that data in a future Albany Data Story.  

Figure 4 (above) - The average percent, by month, of yearly crimes occurring OUTSIDE over the study period, 2020-2024. The increase observed from May-August was more pronounced when OUTSIDE locations were isolated, confirming our earlier hypothesis. The error bars were calculated from the standard distribution of percentage data for any given month. 

A total of 10,922 crimes were classified as OUTSIDE, and Figure 4 shows a more profound increase from May to August before trailing off for the remainder of the year. An average increase of approximately 25% was observed in May-August compared to December-March. This inspired us to further group the data using tags from the UNIFORM CRIME REPORT into three categories: 

THEFT 

• LARCENY, MV LARCENY, STOLEN PROPERTY, BURGLARY, ROBBERY, EMBEZZLEMENT, FRAUD, FORGERY, UNAUTHORIZED USE, EXTORTION

VIOLENCE

• SIMPLE ASSAULT, AGGRAVATED ASSAULT, MURDER, SEX OFFENSE, KIDNAPPING, RAPE, COERCION

OTHER 

• CRIMINAL MISCHIEF, ARSON, CONTROLLED SUBSTANCES, DANGEROUS WEAPONS, OFFENSES AGAINST FAMILY, DISORDERLY CONDUCT, OFFENSES AGAINST PUBLIC ORDER, GCO, DWI, GAMBLING, POSS OF BURGLARY TOOLS, PROSTITUTION. 

The purpose of this division was to separate crimes into two broad groups that reflect a crime done to an individual (VIOLENCE) or their property (THEFT), and to see if either of these categories showed a similar increase from May to August. The remaining ambiguous tags were lumped together into OTHER. We feel strongly that an in-depth analysis could be made on any of these crime tags, especially those that fall under OTHER; however, at this time, the broad grouping into these categories better helps us examine trends and not draw conclusions based on small sample sizes.   

Figure 5 (above) - The average percent, by month, of yearly crimes categorized as THEFT, VIOLENCE and OTHER.   

By separating crimes initially labeled as OUTSIDE into three categories, we can more easily observe a stark difference in crimes labeled as VIOLENCE from May to August (above 10%) relative to December to March (less than 6%). THEFT remains the most constant throughout the year, but is lower on average for the first three months. The frequency of crimes categorized as OTHER tends to fall between THEFT and VIOLENCE.  In total, 5519 were classified under THEFT, 2188 were classified as VIOLENCE, and 3215 crimes were classified as OTHER. The breakdown by year is provided below. Increases year-over-year are highlighted in red, whereas decreases are highlighted in blue. 

Table 4 (above) - The number of OUTSIDE crimes categorized as THEFT, VIOLENCE and OTHER, by year, with year-over-year change (%).   

We consider this to be one of the most interesting findings from our study.  The peak number of THEFT crimes occurred in 2022, and has subsequently declined in both 2023 and 2024 by 13% each year, but is still higher than in 2020 and 2021. VIOLENCE crimes have generally risen each year since 2020, with 2022 being the one year showing decline.