What multivariate analyses could be performed with employment tenure and temporary employment rate before and after labour reforms?

[Eduard López]

I'm studying the employment tenure (the duration of contracts) and the temporary employment rate of certain social groups before and after labour market reforms. So far I've performed descriptive analyses, but I would like to consider doing some multivariate analyses of the variables so I can do some causal inference beyond "looks like employment tenure has gone down for female workers, perhaps they were affected by the reform".

My question is, what sort of multivariate analyses can be performed with a continuous variable such as employment tenure and with the temporary employment rate? If I understand it right, the independent variable would be the year "after the reform" (as the event), but how can I measure the (purported) effect on employment tenure or the temporary employment rate?
Thanks a lot!

[Jared Greathouse]

My recommendation, without knowing the details or specifics, would be difference-in-differences, synthetic controls, or interrupted time series. The independent variable is the policy, and the causal effect is the difference between the observed outcomes and what the outcomes of the treated units WOULD'VE been had they not been treated.

[Eduard López]

Thanks a lot Jared, are there any resources you would recommend me to look up to learn how to follow the differences in differences, synthetic controls or interrupted time series?

My idea was to measure if the temporary employment rate or the median duration of labour contracts had experienced any change, to I was hoping to make them the DV, and the independent variable would be the policy (or in my case, the year, as I would be studying the year before the policy was implemented and after, as a proxy for the policy).

[Jared Greathouse]

Plenty of resources, but the most digestible are this book, this book, and this article on ITS. The other two books cover many introductory causal designs.