The Bottom Line: Violent crime on O‘ahu is currently: mostly stable.
Each month, we analyze the crime data provided publicly by the Honolulu Police Department to understand crime trends and put the headlines in perspective. Currently, for each category of violent crime, we compare the number of crimes last month to the month before and the same month in the previous year. We do a simple statistical analysis to determine whether the year-over-year or month-to-month variation is out of the norm or expected variation.
Additionally, we are beginning to offer some longer-term analyses using more data, not just our standard quick assessment. Using data we have compiled going back to July 2024, and August through November of 2023, we can say more about the distribution of crime rates within each category and also speak to longer-term trends. One caveat to these analyses is we use the numbers we have reported on our blog, which rely on the HPD numbers reported at the time when we wrote each post; consequently, our numbers are frozen from the time of the blog post whereas the HPD numbers fluctuate slightly as they are updated. As noted in our October post, we decided that rather than updating all of our numbers each month, we will use the previously posted numbers; while this results in the numbers being off typically by a few counts (about 1-6), it should not affect the statistical significance in most categories.
If this is your first visit to the Crime Lab, check out our primer on understanding crime rates.
Aggravated Assault: Down/Stable (And Within the Distribution for July to December)
Quick Assessment: There were 77 aggravated assaults in December. This is down from December 2023 when there were 92. It is also down from the number in November when there were 84.
Is this change significant? Actually, some of them. While the year-over-year decrease is statistically significant, the month-to-month decrease is not quite statistically significant.
Longer-term Assessment: After a spike in aggravated assaults in October, the number of aggravated assaults in November and December returned to be “in the distribution” or the normal amount of aggravated assaults, which have ranged between 72 and 92 since July (excluding October, which was out of the typical distribution). Moreover, despite the peak in October, so far, we’re at about the same average number of aggravated assaults for August through December (89.8 for those months in 2023 and 85 for those months this year).
Murder and Non-Negligent Homicide: Stable (And Within the Distribution for July to December)
Quick Assessment: There were 2 homicides in December. This is down slightly from December 2023 when there were 4. It is up from November when there were 0.
Is this change significant? No. Neither the year-over-year drop nor the month-to-month increase are statistically significant, meaning they are within expected variation.
Longer-term Assessment: Homicides this year since July have varied from 0 to 6; in part because these are generally low numbers (of awful crimes), they tend to be very “noisy” meaning there is a lot of fluctuation. Reports in percentages can make the changes seem more meaningful than they are. During the second half of 2024, however, our homicide rate has been fairly stable—excepting August, there were 0 or 2 homicides each month—and the homicide rate is currently within the expected distribution. Our average between August and December for 2024 is about the same as in 2023 (2.4 in 2024 and 2.8 in 2023).
Robbery: Stable (And Within the Distribution for July to December)
Quick Assessment: There were 48 robberies in December. This is down from December 2023 when there were 55. It is also down from the number in November when there were 61.
Is this change significant? Actually, no. Neither the year-over-year decrease nor the month-to-month decrease are statistically significant, meaning December’s number is within the expected variation.
Longer-term Assessment: Robberies ticked up slightly in October and November, but they are back to the lower numbers we saw in July and September. In 2024, since July, they have ranged from 38 to 73, so at 48 we’re in the low end/middle of that distribution. Our average between August and December for 2024 is about the same as in 2023 (55.6 this year and 51.2 last year). In the final six months of 2024, however, our numbers fluctuated more than they did in these same months in 2023.
Sexual Assault: Up/Stable (And Within the Distribution for July to December)
Quick Assessment: There were 74 sexual assaults in December. This is up from December 2023 when there were 46. It is up slightly from the number in November when there were 66.
Is this change significant? Actually, some of them are. While the month-to-month increase is not statistically significant, the year-over-year increase is significant.
Longer-term Assessment: Sexual assaults this month continue to be within the typical distribution. In 2024, since July, they have ranged from 64 to 79, so at 74 we’re at the high end of that distribution. Likewise, our average between August and December 2024 is about the same as in 2023 (70.8 this year and 64.6 last year). October of 2023 saw a particularly bad spike in the number of sexual assaults (93) that thankfully we did not see this year, but that was followed by a much lower number the following month (56). While sexual assaults lowered substantially in the close of last year, we did not see a similar decline in 2024, with sexual assaults staying fairly level over the final six months of 2024.
Next Steps
The Hawai‘i Crime Lab officially launches this month so undergraduate students from UHM will be assisting with data collection so we can construct graphs to show what the trends look over time like instead of providing snapshots. In late January, we hope to be able to show some full-year plots comparing 2023 and 2024; in future months’ posts, we can show where that month’s rate is relative to where we were at the same time in prior years. As always, we hope to continue to pull back the curtain and discuss how we calculate the expected variation. For now, we hope our analysis provides some context.
Still on our to do list as well is to provide another special look at crime in West O‘ahu as a followup to our last post putting the crime trends in statistical perspective. Again our goal there is to see to what extent trends look different in West O‘ahu since the numbers we analyze are island wide and may be missing more area-specific variations and trends.
Final Assessment
This month’s assessment is that violent crime on O‘ahu is mostly stable. This month’s results continue to remind us that fluctuations are to be expected and we really want to look at trends.
Although it is too soon (given the data we have collected—six months in 2024 and 5 in 2023) to say if the crime rate is going up or down (what people generally want to know), we can say something about the level of noise in the data. So far, one clear trend is that sexual assaults are staying consistent but consistently higher than they were in the same months in 2023 (although not significantly so). Likewise, the homicide rate has been far more stable in 2024 than in 2023. By contrast, the robbery rate is showing more volatility in 2024 than in the same months in 2023, as has the number of aggravated assaults.
The Hawaii Crime Lab 