For my applied microeconomics course this semester we need to set up a hypothetical test for an outcome variable to show the impact that covid-19 has had on the outcome variable. I would like to test employment inequality between Indigenous and non-Indigenous Australians. This is my idea so far: build an estimation model using historical data ( say the past 3 censuses (2006-2016)) with variables such as individual characteristics (ie age, male/female, married, education etc.) household characteristics (ie household size, mixed race household) and area characteristics (ie remoteness etc). Then using the values for all these variables from the 2020 census coming up with an estimate for employment. This estimate (for Indigenous and non-Indigenous) would effectively be the employment rate had covid-19 not occurred. Compare this to the actual employment rate and you have the impact of covid. Is this too presumptuous? Does anyone have a suggestion for improvements or a completely different idea? Worth noting is the fact that employment levels for both groups have barely changed over the past 10-15 years. (approximately 50% for Indigenous and 75% for non-Indigenous)
