Interpreting results after taking first differences

[mjamei]

Hi,

I had a question about how I should go about interpreting my results. I've taken first differences of my dependent and independent variables and used a model to estimate inflation expectations (my dependent variable) but I'm unsure what to do next. My main problem is that my dependent variable as well as some of my independent ones are percentage rates (like inflation rates, unemployment rate etc), so I don't know how taking first differences translates into interpretations.

I was told by my tutor that after taking first differences, results are interpreted by per cent increases rather than unit increases but it doesn't really make sense to take percentage increases of a variable which is already in per cent format. I read somewhere else that you can interpret your results as percentage point differences, which would make more sense for my results, but goes against what my tutor told me.

Can anyone shed some light on this please?

Thanks.

[qwerty87]

The important thing is to remain as consistent as possible, if one of your variables is in percentage then all others should also be percentage. Second, why are you taking first differences? usually the reason is that you have non-stationary variables. For interpretation on the other hand, you shouldn't worry about unit increase or percentage increase, you should on the other hand hypothesize firsthand about your independent variable, such as its sign and its relationship with the dependent variable. You do not need to interpret first differences as something else, that is if you have a good reason for taking first differences.