Skip to contents

Detect surges in stop and search activity

Source: R/treatment-define.R
lamp_detect_surges.Rd

Flags area-months in which searches rise far above the area's recent level, and returns them as a candidate staggered-adoption table. Two methods are offered: a threshold of k standard deviations above the area's trailing twelve-month mean, and a structural break test from strucchange.

Usage

lamp_detect_surges(
  panel,
  method = c("threshold", "breakpoints"),
  k = 2,
  min_months = 12L,
  min_stops = 3L
)

Arguments

panel

A lamp_panel with a stops column.

method

"threshold" (default) or "breakpoints".

k

Threshold in standard deviations above the trailing mean.

min_months

Months of history required before an area can be flagged.

min_stops

Minimum searches in the month, so that tiny areas do not surge from one search to three.

Value

A tibble of class lamp_surges with one row per area: area, adoption_month (the first flagged month, NA if none), n_flagged, trailing_mean, trailing_sd, peak_stops and peak_ratio (peak over trailing mean). Pass it to lamp_treatment() as adoption for a staggered design. The attribute flags holds every flagged area-month.

Endogeneity

A surge detected from the stop series is not an experiment. Police deploy searches where crime has risen, so an area's first surge month is correlated with its recent crime history and with anything else that prompted the deployment. Treating a detected surge as exogenous will mistake the response for the cause. Use this to explore where activity changed and to pick candidate cases for a design with an external source of variation, not as the primary design; and read lamp_allocation() alongside any estimate that uses it.

See also

Other treatment: lamp_treatment()

Examples

sim <- lamp_simulate(n_areas = 16, n_months = 30, design = "event", seed = 3)
surges <- lamp_detect_surges(sim, min_stops = 5)
head(surges)
#> streetlamp surges: 6 of 6 areas flagged (threshold, k = 2).
#> ! Surges detected from the stop series are endogenous; see `lamp_detect_surges()`.
#> # A tibble: 6 × 7
#>   area  adoption_month n_flagged trailing_mean trailing_sd peak_stops peak_ratio
#>   <chr> <date>             <int>         <dbl>       <dbl>      <dbl>      <dbl>
#> 1 S000… 2019-04-01             3          7.79        3.48         16       2.05
#> 2 S000… 2019-03-01             2          7.80        3.46         16       2.05
#> 3 S000… 2019-03-01             2          8.24        3.75         16       1.94
#> 4 S000… 2019-03-01             2          8.90        3.49         16       1.80
#> 5 S000… 2019-03-01             5          8.40        4.30         21       2.50
#> 6 S000… 2019-03-01             1          9.32        4.18         19       2.04