Where Should the Overtime Patrol Hours Go?
After a rise in thefts from vehicles, a divisional commander asks you to compare three districts for the next quarter. Reported theft-from-vehicle cases in the four quarters of last year were: Kwun Tong 120, 150, 180 and 210; Wong Tai Sin 80, 75, 70 and 60; Sai Kung 40, 55, 70 and 95. Populations are 650,000, 420,000 and 480,000 respectively. Public car-parking spaces are 8,000 in Kwun Tong, 4,500 in Wong Tai Sin and 6,000 in Sai Kung. The commander has 1,200 officer-hours of overtime. One foot-patrol shift uses two officers for four hours. The commander says, 'Kwun Tong has the biggest numbers, so put everything there.' You are asked to test that instruction against rates, trends and operational practicality before the deployment meeting tomorrow.
1. Calculate the annual theft-from-vehicle rate per 100,000 population and per 1,000 public parking spaces for each district. Show your working and rank the districts on each measure.
Read the worked answer
Totals: Kwun Tong = 120+150+180+210 = 660; Wong Tai Sin = 80+75+70+60 = 285; Sai Kung = 40+55+70+95 = 260.
Per 100,000 population:
- Kwun Tong: 660 ÷ 650,000 × 100,000 = 101.5
- Wong Tai Sin: 285 ÷ 420,000 × 100,000 = 67.9
- Sai Kung: 260 ÷ 480,000 × 100,000 = 54.2 Rank: Kwun Tong > Wong Tai Sin > Sai Kung.
Per 1,000 parking spaces:
- Kwun Tong: 660 ÷ 8,000 × 1,000 = 82.5
- Wong Tai Sin: 285 ÷ 4,500 × 1,000 = 63.3
- Sai Kung: 260 ÷ 6,000 × 1,000 = 43.3 Rank: Kwun Tong > Wong Tai Sin > Sai Kung.
The two denominators tell a consistent story at annual level, but they answer different questions: population rate reflects resident risk; parking-space rate reflects opportunities at public car parks. The parking-space measure is more directly actionable for foot patrol, while the population measure is better for comparing the burden on residents.
2. Allocate the 1,200 officer-hours across the three districts. Show the number of shifts, the hours and a reason for each allocation. Identify one operational risk.
Read the worked answer
One shift = 2 officers × 4 hours = 8 officer-hours. 1,200 ÷ 8 = 150 shifts. A defensible allocation: Kwun Tong 70 shifts (560 hours), Sai Kung 50 shifts (400 hours), Wong Tai Sin 30 shifts (240 hours). Total = 150 shifts / 1,200 hours.
Reasons: Kwun Tong has the highest rate on both measures and the largest absolute burden (660 cases); it needs the core allocation. Sai Kung has the fastest upward trend (40 to 95, +137.5%) and by Q4 has 95 cases, so a purely annual ranking would miss the emerging problem; 50 shifts buys visible presence in the highest-growth car parks. Wong Tai Sin is falling (80 to 60, -25%) but still has 285 cases; 30 shifts maintains a baseline and avoids abandoning a district that may be under-reporting.
Risk: displacement - heavy patrols in Kwun Tong may push offenders to Wong Tai Sin or to private car parks; the allocation should be reviewed after four to six weeks with weekly theft data, not fixed for the whole quarter. A second risk is overtime fatigue and reduced response capacity elsewhere.
3. The commander says: 'Kwun Tong has the biggest numbers, so put everything there.' What data limitations and decision errors should you point out?
Read the worked answer
First, 'biggest numbers' is not the same as highest risk: Kwun Tong also has the largest population and the most parking spaces, so its absolute count is expected to be larger. Rates and trends are more comparable. Second, reported cases depend on reporting behaviour; a hotline change or a local awareness campaign can move counts without changing theft. Third, population is a residential denominator, but theft from vehicles may be committed by people who drive into the district; daytime or visitor population would be better. Fourth, parking spaces are not equal in usage, lighting, CCTV or turnover; a space in a busy mall car park is not the same as a quiet lay-by. Fifth, the data are annual totals and hide seasonality; Q4 in Sai Kung may be a seasonal spike or the start of a trend. Sixth, displacement can make the district look better while total crime is unchanged. Putting everything in one district would also create gaps elsewhere and risk officer fatigue. Recommend a shorter review cycle and a small comparison area.