What we should be controling for in our analysis?

[vob2]

Economy Industry Firm characteristics that change over time and might affect project delays

[harishk05]

The Barrot and Nanda paper provides several robustness tests in their appendix. Some of these may be relevant for our paper as well.

Two other thoughts:

• Could quick payment affect the composition of firms (small business primes)? In the original (non-quickpay) world, "weak" small businesses may not bid for government contracts because of the late payment. (Weak firms could be poorly managed firms that do not do a good job of managing their working capital; and this may be correlated with weak management overall.) Under quickpay, some weak firms may bid and win some contracts. The increase in delays may then not be a direct (moral hazard) effect on the behavior of firms, but may be the result of more weak firms entering the government contracting world.

• The same argument can be made as a survival effect; more weak firms may survive under quickpay.

The way to address this may be to only consider firms that exist both pre- and post-quickpay.

[vob2]

Thank you, Harish. Any specific tests in Barrot and Nanda or any priority on what we should try first?

[harishk05]

Not sure Vlad. I will have to look at it more closely. Footnote 31 (page 23) in their paper (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2808666) lists various robustness tests. Not sure which may be relevant to us. Some of those tests seem relevant to their application (effect on employment), but some may be relevant to us. Let me look at it more closely and try to get back.

[vibhuti6]

Hi Vlad and Harish, I posted some new results for quarter-to-quarter delays here.

This page includes 1) verification of parallel trends, 2) controls (fixed effects) for firm, task, and industry, and 3) subsample analyses for firms and contracts that exist in both pre- and post- quickpay period.

Some follow up thoughts/comments:

1) At this point, it is not clear to me how we can control for changes to the economy over time in our analysis. I need to think more about this issue.

2) Regarding possible changes to the composition of small firms: I think this is a good test for evaluating the mechanism driving the treatment effect. As such, the results remain robust for this subsample. In general, my understanding is that change within small businesses over time should not be a threat to our identification strategy. Because we are interested in the interaction effect, i.e. the change in small businesses over time relative to the corresponding change in large businesses.

3) I will go through the appendix in Barrot and Nanda paper as well. Thanks.

[vob2]

Hi all,

Numbers refer to Vibhuti's list:

1. Following up on the comments from Jie about monthly observations. Do we have enough variation in time to include time fixed effects in our DiD? Enough variation in contracts per firm to include firm fixed effects?

2. I think the concern here would be that the story is different if it is the composition of firms that drives the delays in project completion vs. the existing firms not working hard enough. Either way it is an effect of the Quickpay law, but the mechanisms are different.

3. Thanks!

Vlad

[vibhuti6]

Hi Vlad,

There seems to be enough variation in each quarter. See graph below:

I will check for variation in each month, and get back to you.

For variation in firms, about half of the firms in the sample have only one contract each, and the other half have multiple contracts. See the table below.

contract group	number of firms
One contract	11444
1-5 contracts	7494
5-10 contracts	2104
10-15 contracts	976
More than 15 contracts	2726

[vob2]

Use the same controls as in Barrot and Nanda. Vibhuti and Jie already have them. Closing issue