How to account for waves of policy

[schuylerfried]

Ex. x county implements policy y. Two weeks later, they remove policy y. Two weeks after that, they re-implement policy y. This type of policy "dance" seems likely after the curve has been flattened. However, it doesn't look like the current schema can capture that. I think it would be useful to have a schema that looks like:

• date, policy, county, state, fips

Data would look like:

--------------------------------------------
4/2/2020, SD, alameda county, CA, 09423
4/2/2020, GS_XX, alameda county, CA, 09423
...
--------------------------------------------

and be updated daily. An alternative is to create a boolean column for each policy. That would mean one new row per county per day.

A second table could map policy abbreviations to actual names

[jadraketx]

Thanks for assembling this resource! Can we expect an update coming soon?

[padelson]

Hey

Thanks for assembling this resource! Can we expect an update coming soon?

This Friday, we're preparing to issue an update expanding this dataset to all 50 states + ~300 counties (all counties with at least 100 confirmed cases as of 4/06/2020).

[padelson]

Ex. x county implements policy y. Two weeks later, they remove policy y. Two weeks after that, they re-implement policy y. This type of policy "dance" seems likely after the curve has been flattened. However, it doesn't look like the current schema can capture that. I think it would be useful to have a schema that looks like:

• date, policy, county, state, fips

Data would look like:

--------------------------------------------
4/2/2020, SD, alameda county, CA, 09423
4/2/2020, GS_XX, alameda county, CA, 09423
...
--------------------------------------------

and be updated daily. An alternative is to create a boolean column for each policy. That would mean one new row per county per day.

A second table could map policy abbreviations to actual names

Thanks for your comment. You're right about the likely rolling nature of these policies. Consequently, going forward we'll report data primarily in the format currently presented as 'raw'. This will include the start / end dates of each policy. As a policy starts / ends multiple times, there are will be multiple rows for that policy. (https://medium.com/@tomaspueyo/coronavirus-the-hammer-and-the-dance-be9337092b56)

[jadraketx]

Will the update be pushed to this repository? Just want to make sure I don't miss it. Thanks!

This Friday, we're preparing to issue an update expanding this dataset to all 50 states + ~300 counties (all counties with at least 100 confirmed cases as of 4/06/2020).

[padelson]

Hey there - just pushed a partial version of the update, with counties + state policies separated. So this data doesn't include state-level policies' impact on the county level.. We'll be pushing another version of the data, where counties inherit state policies, tomorrow morning.