Description

QUESTION 2. Crime and Police [15 marks]

Imagine you start a job as a research assistant in the UK Ministry

of Justice. Your boss asks you to analyse the causal effect of the

number of police officers in the ground on the level of crime in the

streets. Originally, you are given a dataset with monthly data of

number of police officers at the locality level, average education of

adult population in the locality, a wealth index at the local level, as

well as crime data on knife crime, robbery, and homicides.

a) Write down the empirical model that you would consider using

cross-sectional data. [2 mark] What would you do with this

data to explore your main question of interest? [3 mark]

b) What sort of limitations you foresee given the type of the data

you have received? How would you try to solve such problems?

[5 marks]

c) Now imagine you have a panel dataset, could you solve some

of the problems you have mentioned in 2.b using this data? If

so, explain how. Do you think you are still having limitations

to claim your coefficient of number of police officers is causal?

Explain why you may or may not continue having limitations

to give such interpretation. [5 marks]