Dummy Variable Components Interpretation

[sanseveroj]

Hello I work on contact forecasts into a call center which is dependent on holiday factors.

Typically I have represented these as dummy variables to the ARIMA function which works fine.

When I input these variables into FASSTER they work fine but I am confused by the interpretation of components. They show that despite having a 0 value they are having an influence over the target value? Is this correct?

Shouldn't I expect to see just a decrease or increase on the days with a 1 in the dummy variable?

Sorry for the limited information trying to avoid sharing the underlying proprietary data.