This data analysis project is dedicated to discovering, analyzing, and visualizing crime data in Spokane, WA and around the state. Informed residents can make better decisions about where to live, raise families, work, and play in our wonderful city with up-to-date information on crime. With data in-hand, residents can engage actively and factually with local leaders and law enforcement. Using Computer Statistics (CompStat) data provided by Spokane Police Department (SPD) we're able to analyze crime that has already happened, discover trends in the data, and make predictions.
Currently, the SPD makes weekly CompStat reports available each Monday. Reports cover reported crime for the prior week in PDF format, broken down by policing districts. We scrape the PDF files, extract data, and transform it into a format useful for analysis and visualization.
- Overall Statistics
- Seasonality
- Police District Statistics
- Types of Offenses
- Lunar Phases
- Data Sources
- Data Limitations
- Violence Classifications
- Contact
Looking at overall statistics gives us a good starting point from which to explore crime over time and specific types of incidents. From September 2017 onward we see an overall downward trend in reported crime:
We have CompStat data for full years beginning in 2018 for comparison:
| Year | Total Incidents | Percentage Change | Per Capita (100k) |
|---|---|---|---|
| 2018 | 17418 | NA | 7591 |
| 2019 | 14434 | -17.13 | 6291 |
| 2020 | 13321 | -7.71 | 5806 |
| 2021 | 12125 | -8.98 | 5284 |
| 2022 | 12627 | 4.14 | 5503 |
| 2023 | 13178 | 4.36 | 5743 |
| 2024 | 7664 | NA | NA |
While CompStat data does exist prior to 2018, the reporting standards changed and some reports are missing. I'm not able to make accurate comparisons with years prior to 2018.
Total offenses by year and month:
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 1498 | 1235 | 1163 | 1077 | 1087 | 1132 | 976 |
| 02 | NA | 1186 | 916 | 1100 | 831 | 1177 | 1042 | 980 |
| 03 | NA | 1311 | 1080 | 1008 | 950 | 1195 | 969 | 1053 |
| 04 | NA | 1309 | 1305 | 1045 | 923 | 1101 | 984 | 949 |
| 05 | NA | 1524 | 1423 | 1044 | 991 | 1120 | 1190 | 1031 |
| 06 | NA | 1538 | 1312 | 1109 | 1011 | 1009 | 1207 | 942 |
| 07 | NA | 1656 | 1358 | 1205 | 1025 | 860 | 1217 | 1110 |
| 08 | 1277 | 1558 | 1300 | 1243 | 1114 | 910 | 1245 | 623 |
| 09 | 1290 | 1449 | 1226 | 1181 | 1151 | 891 | 1126 | NA |
| 10 | 1376 | 1543 | 1099 | 1209 | 959 | 1147 | 1010 | NA |
| 11 | 1418 | 1479 | 1049 | 1013 | 1010 | 1082 | 1011 | NA |
| 12 | 1454 | 1367 | 1131 | 1001 | 1083 | 1048 | 1045 | NA |
Here are the average daily incidents by month:
The majority of crime in Spokane is non-violent, though assaults, robbery, and other violent crimes do occur with some regularity:
| ARSON | ASSAULT | BURGLARY | DEFRAUD PUBLIC UTILITY | DRIVE-BY | HARASSMENT | HOMICIDE | INTIMIDATE | MANSLAUGHTER | MURDER | POISON | RAPE | ROBBERY | TAKING VEH. | THEFT | VEH. THEFT | VEH. PROWL | VEH. TRESPASS | VEHICULAR ASSAULT | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2018 | 41 | 853 | 2150 | 0 | 18 | 63 | 1 | 40 | 0 | 8 | 0 | 301 | 310 | 70 | 11649 | 1623 | 258 | 3 | 30 |
| 2019 | 42 | 729 | 1858 | 0 | 19 | 70 | 0 | 4 | 0 | 5 | 0 | 217 | 290 | 52 | 9912 | 1205 | 10 | 0 | 21 |
| 2020 | 99 | 663 | 1949 | 0 | 26 | 52 | 0 | 0 | 1 | 10 | 1 | 190 | 266 | 31 | 8937 | 1072 | 0 | 0 | 24 |
| 2021 | 101 | 713 | 1699 | 0 | 30 | 64 | 0 | 0 | 0 | 10 | 1 | 219 | 260 | 63 | 7865 | 1081 | 1 | 0 | 18 |
| 2022 | 58 | 701 | 1658 | 0 | 57 | 68 | 0 | 1 | 0 | 14 | 0 | 192 | 275 | 77 | 7889 | 1610 | 0 | 0 | 27 |
| 2023 | 82 | 750 | 1825 | 1 | 41 | 104 | 0 | 1 | 0 | 9 | 1 | 202 | 304 | 55 | 8332 | 1447 | 0 | 0 | 24 |
| 2024 | 76 | 457 | 918 | 0 | 23 | 90 | 0 | 0 | 0 | 7 | 1 | 123 | 205 | 30 | 4990 | 731 | 0 | 0 | 12 |
Here's both types of crime shown over time:
Of non-violent crimes, theft is the most common. This includes shoplifting, burglary (residential and commercial), and motor vehicle thefts.
When we're looking at events as they happen over time we must consider seasonality. Crime in Spokane definitely follows seasonal trends occurring throughout the year. We can use this seasonality to make predictions and give context to increases or decreases in crime. Data that follows a seasonal trend can be broken down ("decomposed") into its consituent parts. Here's a decomposition of overall crime data:
The "seasonal" section of the plot above shows an obvious, repeating cycle that occurs each year. There's also a downward trend line. Once we've identified seasonal cycles it's possible to visualize what crime is doing independent of the expected change:
The city is divided among eight policing districts, with four (P1, P2, P3, P4) in the "North Police Service Area", and the other four (P5, P6, P7, P8) in the "South". The north and south service areas are separated geographically by the Spokane River. Some offenses are noted as occurring in an "OTH" district with no amplifying information given. Districts designated "SPA", "SPB", "SPC", and "SPD" had also been used in the past but no information is given in the CompStat reports indicating what areas represent. Sometimes, an offense will be listed with no district in the weekly reports, possibly due to a data entry error. There was a spike in missing district information beginning in late 2021 and continuing into 2022.
Here's a graph of offenses with missing district information:
In fact, there are more "NA" reports in 2022 than any other district:
With that much data missing I'm not able to accurately visualize the count of crimes per-district for late 2021 or 2022. I've emailed SPD about this missing data but have yet to hear back. From September 2017 onward, districts P1-4 have the highest overall counts of reported crimes:
Most crime in Spokane is non-violent. These offenses include burglary, motor vehicle theft, and "ordinary" theft.
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 1007 | 817 | 769 | 761 | 693 | 683 | 634 |
| 02 | NA | 783 | 616 | 755 | 528 | 793 | 683 | 651 |
| 03 | NA | 902 | 736 | 662 | 618 | 794 | 606 | 701 |
| 04 | NA | 901 | 919 | 675 | 601 | 736 | 638 | 606 |
| 05 | NA | 1006 | 993 | 693 | 642 | 722 | 752 | 655 |
| 06 | NA | 1042 | 898 | 721 | 684 | 608 | 752 | 601 |
| 07 | NA | 1118 | 924 | 787 | 654 | 479 | 747 | 726 |
| 08 | 918 | 1006 | 906 | 856 | 718 | 471 | 710 | 416 |
| 09 | 862 | 977 | 853 | 820 | 703 | 533 | 763 | NA |
| 10 | 935 | 1029 | 756 | 856 | 578 | 685 | 654 | NA |
| 11 | 932 | 1017 | 715 | 648 | 642 | 694 | 668 | NA |
| 12 | 985 | 861 | 779 | 695 | 736 | 681 | 676 | NA |
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 175 | 167 | 155 | 146 | 156 | 195 | 136 |
| 02 | NA | 118 | 117 | 141 | 118 | 128 | 121 | 115 |
| 03 | NA | 128 | 135 | 175 | 143 | 124 | 131 | 105 |
| 04 | NA | 173 | 171 | 160 | 117 | 129 | 132 | 119 |
| 05 | NA | 177 | 185 | 157 | 134 | 153 | 162 | 131 |
| 06 | NA | 171 | 170 | 166 | 146 | 154 | 163 | 104 |
| 07 | NA | 197 | 179 | 187 | 137 | 114 | 150 | 133 |
| 08 | 148 | 215 | 162 | 149 | 151 | 152 | 161 | 75 |
| 09 | 153 | 193 | 154 | 171 | 169 | 127 | 149 | NA |
| 10 | 159 | 203 | 146 | 172 | 148 | 151 | 141 | NA |
| 11 | 184 | 189 | 137 | 162 | 154 | 144 | 148 | NA |
| 12 | 160 | 211 | 135 | 154 | 136 | 126 | 172 | NA |
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 165 | 91 | 127 | 81 | 129 | 131 | 90 |
| 02 | NA | 133 | 93 | 107 | 66 | 148 | 105 | 85 |
| 03 | NA | 120 | 100 | 81 | 56 | 125 | 113 | 111 |
| 04 | NA | 96 | 91 | 97 | 72 | 119 | 89 | 97 |
| 05 | NA | 140 | 115 | 70 | 97 | 119 | 139 | 121 |
| 06 | NA | 140 | 116 | 93 | 45 | 112 | 139 | 87 |
| 07 | NA | 136 | 117 | 95 | 94 | 146 | 158 | 92 |
| 08 | 82 | 148 | 96 | 88 | 116 | 165 | 227 | 48 |
| 09 | 124 | 113 | 86 | 82 | 133 | 119 | 86 | NA |
| 10 | 123 | 137 | 92 | 73 | 106 | 169 | 86 | NA |
| 11 | 152 | 143 | 92 | 97 | 114 | 126 | 80 | NA |
| 12 | 144 | 152 | 116 | 62 | 101 | 133 | 94 | NA |
Motor vehicle theft data table
Violent crime includes assault (vehicular and otherwise), drive-by shootings, sexual assault, robbery, intimidation with a weapon, and carjacking.
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 26 | 28 | 23 | 13 | 25 | 15 | 27 |
| 02 | NA | 24 | 17 | 22 | 24 | 20 | 24 | 25 |
| 03 | NA | 19 | 21 | 28 | 17 | 31 | 20 | 32 |
| 04 | NA | 15 | 19 | 24 | 26 | 21 | 27 | 26 |
| 05 | NA | 15 | 22 | 17 | 19 | 25 | 31 | 26 |
| 06 | NA | 21 | 24 | 21 | 28 | 30 | 29 | 24 |
| 07 | NA | 41 | 27 | 29 | 22 | 27 | 45 | 30 |
| 08 | 18 | 32 | 28 | 25 | 21 | 16 | 30 | 15 |
| 09 | 21 | 34 | 28 | 23 | 25 | 19 | 22 | NA |
| 10 | 20 | 28 | 25 | 24 | 22 | 23 | 24 | NA |
| 11 | 19 | 23 | 26 | 10 | 24 | 22 | 24 | NA |
| 12 | 34 | 32 | 25 | 20 | 19 | 16 | 13 | NA |
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 59 | 66 | 62 | 43 | 52 | 60 | 48 |
| 02 | NA | 68 | 43 | 45 | 52 | 53 | 61 | 63 |
| 03 | NA | 69 | 52 | 37 | 74 | 63 | 52 | 57 |
| 04 | NA | 66 | 69 | 54 | 53 | 52 | 58 | 51 |
| 05 | NA | 97 | 73 | 70 | 65 | 48 | 70 | 46 |
| 06 | NA | 81 | 64 | 61 | 61 | 65 | 74 | 76 |
| 07 | NA | 82 | 74 | 71 | 75 | 52 | 84 | 71 |
| 08 | 58 | 82 | 67 | 67 | 60 | 72 | 65 | 45 |
| 09 | 62 | 63 | 70 | 46 | 60 | 57 | 57 | NA |
| 10 | 58 | 81 | 48 | 44 | 66 | 77 | 64 | NA |
| 11 | 55 | 49 | 50 | 59 | 43 | 51 | 58 | NA |
| 12 | 65 | 56 | 53 | 47 | 61 | 59 | 47 | NA |
| Month | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|---|---|---|
| 01 | NA | 17 | 27 | 14 | 16 | 17 | 14 | 11 |
| 02 | NA | 19 | 18 | 11 | 19 | 14 | 21 | 14 |
| 03 | NA | 24 | 20 | 11 | 31 | 21 | 17 | 13 |
| 04 | NA | 25 | 18 | 14 | 24 | 18 | 18 | 21 |
| 05 | NA | 28 | 20 | 18 | 15 | 21 | 19 | 21 |
| 06 | NA | 41 | 21 | 22 | 23 | 19 | 20 | 19 |
| 07 | NA | 22 | 18 | 14 | 12 | 15 | 10 | 13 |
| 08 | 11 | 31 | 24 | 20 | 16 | 11 | 17 | 11 |
| 09 | 18 | 30 | 16 | 21 | 26 | 18 | 22 | NA |
| 10 | 32 | 25 | 14 | 20 | 11 | 13 | 15 | NA |
| 11 | 28 | 20 | 12 | 14 | 15 | 13 | 12 | NA |
| 12 | 26 | 19 | 9 | 11 | 11 | 12 | 17 | NA |
Despite popular sayings like, "must be a full moon out, people are being crazy!", the moon's phase has no noticable affect on crime in Spokane. In fact, crime has been lowest in the last few years when the Moon is full:
There is no significant difference in the amount of offenses between lunar phases.
There are many public resources that we leverage for data analysis. All are open and available to the public, though not always easy to find. Our data sources include the following:
- City of Spokane COMPSTAT reports
- City of Spokane OpenData GIS portal
- Spokane County ArcGIS portal
- HealthData.gov
- FBI's Uniform Crime Reporting (UCR) Statistics Tool
- U.S. Bureau of Labor Statistics Data Tools
- FBI's Crime Data Explorer (CDE)
- Spokane County 911 Communications Annual Reports
There are limitations to the analysis that's possible using CompStat data provided by the city. While detailed information about Spokane crime does eventually get reported in the FBI's NBIRS program, there is often a year's lag or more before the public can download it. For example, 2018's data was made available in September 2019. As of January 2022, the most current data available is for 2020. It will likely be late-2022 or sometime in 2023 before 2021's data can be downloaded and analyzed. In the meantime, we only have the less-detailed CompStat data to work with. More in-depth analysis of the detailed FBI data will become a separate project when I have time.
Weekly CompStat data may not accurately reflect the final offense someone is charged with. For example, simple assault may become homicide if the victim later dies from their injuries. Shoplifting could become "Retail Theft With Special Circumstances" if it's found the offender was part of a larger shoplifing ring. CompStat reports are not retroactively updated to reflect the final disposition of an offense. With that being said, there is more than enough data to identify trends over time and make predictions.
In the SPD's CompStat data, locations of sexual assault offenses are not included in order to protect the victim's privacy. While we know how many sexual assaults are reported and the general policing district that they occurred, it's not possible to correlate that type of assault with specific locations.
When locations are included for other offenses, they are sometimes documented only as an intersection of two roads, with no street numbers or direction. If an address is included in CompStat data, the street number is rounded - a robbery at "1215 E. Main Street" may become "1200 E. Main Street". Sometimes low-numbered locations are documented as "0 E. Main Street". This makes density or "heat" mapping difficult, and the resulting graphics unreliable.
There are some distinctions made in the data provided by SPD, and these reflect the definition of certain offenses in the state of Washington. For example, while purse snatching is robbery in that property is being taken from somebody in-person, it's not the same as being robbed at gunpoint or carjacked. Ideally the perpetrator grabs the purse and flees, without brandishing a weapon or harming the victim otherwise.
In the raw data, there are a number of offenses listed as "ROBBERY 2D PERSON (NOT PURSE SNATCHING)". So while the victim's purse was taken in-person and they were "robbed", we won't see it listed as a robbery. This is not so much a shortcoming of the data as it's a nuanced legal distinction. Spokane Police don't categorize offenses, e.g. violent and non-violent, in CompStat reports, so we're left to do that ourselves.
We're missing some data not due to legal details, but because many crimes go unreported. This occurs for a number of reasons depending on the community - social pressures, past negative encounters with law enforcement, prior criminal history of potential reporters, etc. It would be naive to think that all crime committed in Spokane is reported and subsequently documented in the CompStat reports. Assumptions about the percentage of crimes that go unreported depend on complex, interconnected variables that change over time. It'd take a background in statistics, sociology, and criminal justice that I just don't have to develop solid conclusions. For that reason, this project only focuses on analysis and reporting with the data at-hand, without making assumptions about unreported crimes.
Spokane's reporting of crime data has also undergone multiple transformations in the last decade. CompStat data is available on the SpokaneCity.org portal from 2019 all the way back to 2015. On October 4, 2016 the SPD moved from one crime reporting standard (UCR) to another (NIBRS). Current CompStat reports state on the first page about the switchover date,
Numbers on CompStat reports prior to this date should not be used as a comparison to those on this report.
Unfortunately, we're not able to make true "apples-to-apples" comparisons across the full timespan of availalble CompStat crime data. Some years also have missing weeks of reports, and it's unclear if that data will ever be available.
In some of the analysis I've made distinctions between violent and non-violent crimes. For the purposes of this project, violent crimes include assault, drive-by shooting, homicide, robbery, rape, and murder. While this isn't an exhaustive list of all possible violent crime on the books in our state, it fits the data provided by SPD. Each week when new data is processed I verify that offenses are parsed and categorized properly. If additional types of violent crime are committed and reported in the data I'll add them to the list and they'll become part of the classification process.
Thanks for your interest in the project. If you have questions, comments, or press inquiries feel free to email tyler[at]manitonetworks.com.