Consider the following variables. They are measured on different cities. There are 92 observations, thus the models are estimated using data from 92 cities.

crmrte - crimes per 1000 people
unem - unemployment rate in per cent
pcinc - per capita income
popden - people per sq mile
offarea - police officers per sq mile
lawexpc - law enforce. expend. pc, $
polpc - police per 1000 people
We have the data on the above variables, consider the following models

crmrte_i = \beta_1 + \beta_2 unem_i + u_i \tag{1}

crmrte_i = \beta_1 + \beta_2 unem_i + \beta_3 pcinc_i + u_i \tag{2}

crmrte_i = \beta_1 + \beta_2 unem_i + \beta_3 pcinc_i + \beta_4 popden_i + u_i \tag{3}

crmrte_i = \beta_1 + \beta_2 unem_i + \beta_3 pcinc_i + \beta_4 popden_i + \beta_5 polpc_i + u_i \tag{4}

crmrte_i = \gamma_1 + \gamma_2 polpc_i + u_i \tag{5}

crmrte_i = \gamma_1 + \beta_2 unem_i + \gamma_2 polpc_i + u_i \tag{6}