There goes the neighborhood does tipping exist amongst income groups?

The ‘tipping point’ model of segregation Schelling (1971) has rarely been tested for socioeconomic variables besides race. Using US Census data from 1970 to 2010, I find evidence for tipping behavior amongst income groups. This is strongest when the population is split at the 10th percentile of income. In this case, average tipping points range from 7 to 12%. When the proportion of people below the 10th percentile in a neighborhood exceeds said tipping point, the decrease in those above the 10th percentile will be from 5 to 14%. Robustness checks indicate that these effects are not misinterpreted racial tipping points, nor a result of peoples’ income percentile changing over their lifetime. Effects are strongest in neighborhoods with more school-aged children, suggesting that income tipping is driven by concerns that the provision of local public goods could drop from an influx of lower class residents.