Segmented Regression, Threshold Effects, and Police Expenditures in Small Cities
质疑小城市政府支出通常用单一平滑函数表示的做法，认为分段连续更现实。用分段回归检验明尼苏达、北达科他和威斯康星三州小城市人均警察支出，发现存在不连续分段，且人口增减影响结果。
Abstract Individuals projecting local government expenditures in small cities typically assume that average costs can be represented by a single, smooth, continuous function. This paper discusses why assuming piece‐wise continuity may be more realistic. Segmented regression is then used to test for discontinuities in local per capita expenditures for police protection in Minnesota, North Dakota, and Wisconsin. Three distinct segments were found for expenditures in North Dakota and Wisconsin; two for Minnesota. Minnesota and Wisconsin results improved further when cities with declining and nondeclining populations were treated separately.