race <- rbinom (10000, 1, 0.5)
income <- rnorm (10000, 20000, 10000)*(race+1)
wealth <- rnorm (10000, 20000, 10000)*(race+1)
income <-scale(income)
wealth <- scale (wealth)
crime <- rnorm(10000, 0, 1) - income - wealth

data<-as.data.frame(cbind(race, income, wealth, crime))
colnames(data)<-c("race", "income", "wealth", "crime")

library(dplyr)
data$incomeTier <- case_when((data$income < -1) ~ 1,
          (data$income < 0) & (data$income > -1) ~ 2,
          (data$income < 1) & (data$income > 0) ~ 3,
          (data$income < 2) & (data$income > 1) ~ 4,
          (data$income > 1) ~ 5
          )

data$wealthTier <- case_when((data$wealth < -1) ~ 1,
          (data$wealth < 0) & (data$wealth > -1) ~ 2,
          (data$wealth < 1) & (data$wealth > 0) ~ 3,
          (data$wealth < 2) & (data$wealth > 1) ~ 4,
          (data$wealth > 1) ~ 5
          )

dataWhites<- data[ which(data$race==1), ]
dataBlacks<- data[ which(data$race==0), ]

aggregate(dataWhites$crime, by=list(dataWhites$incomeTier), FUN=mean)[2]
aggregate(dataBlacks$crime, by=list(dataBlacks$incomeTier), FUN=mean)[2]