The EPS currently uses the most recently-available data on union representation percentage disaggregated by ISIC code. For each ISIC code, this percentage is constant over the model run. However, union representation share of many industries has been declining for many years. In the BAU case, we could use a declining rate of union representation for each ISIC codes, based on extrapolating historical data. Also, we should add a lever (policy lever or control setting) that increases union membership, thereby reversing the declining trend. For example, a BAU setting might reduce the union membership percentage by 1.4% per year (approaching an asymptote at zero), but the policy lever would allow for this to be zeroed out or reversed, approaching an asymptote at 1.
This would affect the "Change in Union vs. Non-Union Jobs" output graph but would probably have no effect on any other model output.
The EPS currently uses the most recently-available data on union representation percentage disaggregated by ISIC code. For each ISIC code, this percentage is constant over the model run. However, union representation share of many industries has been declining for many years. In the BAU case, we could use a declining rate of union representation for each ISIC codes, based on extrapolating historical data. Also, we should add a lever (policy lever or control setting) that increases union membership, thereby reversing the declining trend. For example, a BAU setting might reduce the union membership percentage by 1.4% per year (approaching an asymptote at zero), but the policy lever would allow for this to be zeroed out or reversed, approaching an asymptote at 1.
This would affect the "Change in Union vs. Non-Union Jobs" output graph but would probably have no effect on any other model output.
Suggested by Betony Jones on 7/10/2020