Listening to the presentation, I had a question (as an audience member):
When do delays happen? At the beginning or the end of a project? If we divide a project's timeline into triads, 1st, 2nd, 3rd half of the project (CarTalk reference, sorry if you've never heard this NPR show), what percentage of total delays happen in each one?
I guess that most delays happen at the end of the project, but maybe my intuition is wrong, in which case this could be an interesting observation.
The reason this is important is that projects in complexity and predictability and firms differ in their abilities to estimate the completion time. Granted we have project type and firm fixed effects, so we are already capturing some of that.
Perhaps delays after the 2011 implementation went up because a lot of projects of small firms were already at the end of their lives. So this is coincidental with the law. This might be a useful control to add for robustness.
Listening to the presentation, I had a question (as an audience member):
When do delays happen? At the beginning or the end of a project? If we divide a project's timeline into triads, 1st, 2nd, 3rd half of the project (CarTalk reference, sorry if you've never heard this NPR show), what percentage of total delays happen in each one?
I guess that most delays happen at the end of the project, but maybe my intuition is wrong, in which case this could be an interesting observation.
The reason this is important is that projects in complexity and predictability and firms differ in their abilities to estimate the completion time. Granted we have project type and firm fixed effects, so we are already capturing some of that.
Perhaps delays after the 2011 implementation went up because a lot of projects of small firms were already at the end of their lives. So this is coincidental with the law. This might be a useful control to add for robustness.