Questions on How To Count Workers

Hi Folks -

How's everyone doing? Kind of a tough time these days.

To that end, at work I've been working on something to try to understand the impact of COVID-19 closures on Massachusetts' most vulnerable workers. So far we've looked at part-time vs. full-time workers (workers who've worked 48+ wks and 35+ hrs per week vs. everyone else), and gig-economy (ish) workers (who we've set to class of worker == 7, self-employed, unincorporated).

But I'm not confident we're using the best variables to understand these problems. So I'm wondering if folks on the forum have some ideas about how to use PUMS data to obtain these numbers for NAICSP job codes:

* Part-time vs. full-time employment by NAICSP

* Gig-Economy workers

* Tipped Workers

* Undocumented Workers (this has been particularly challenging)

I'm relatively fluent in the coding and aggregating, I'm just having trouble figuring out some best practices towards getting at these numbers. If anyone had any suggestions, I'd be delighted to hear from you.

Thanks,

Peter

  • Hi Peter. Since you are already familiar with the COW class of worker) and NAICSP fields, I will just comment on other important fields for classifying part-time and full-time workers. ESR (Employment Status Recode) can help you zero in on those who were actually employed at time of interview (values 1 and 2). Then, there is a group of variables that can help you distinguish full-time vs part-time workers: WKHP - usual number of hours worked per week; WKL - when last worked; WKW - weeks worked in the past 12 months; WRK - worked last week (yes or no). BTW, ESR can also be used to find those in the active miitary and distinguish them from the civilian workers. Since the ACS never asks if a person is in the U.S. legally, you would have to make some inferences about the people whose records show not a citizen for the CIT variable. (There are researchers who have developed ways to estimate the number of undocumented workers, but I don't know how that is done - Google about. ) I would not recommend using a value of 7 for the COW variable to find gig-economy workers since people with small businesses are also covered by that COW value. So, I don't really know any way to isolate gig-economy workers in the ACS. Perhaps others can chime in with their thoughts on this.

    Doug Hillmer
  • Hi Peter. Since you are already familiar with the COW class of worker) and NAICSP fields, I will just comment on other important fields for classifying part-time and full-time workers. ESR (Employment Status Recode) can help you zero in on those who were actually employed at time of interview (values 1 and 2). Then, there is a group of variables that can help you distinguish full-time vs part-time workers: WKHP - usual number of hours worked per week; WKL - when last worked; WKW - weeks worked in the past 12 months; WRK - worked last week (yes or no). BTW, ESR can also be used to find those in the active miitary and distinguish them from the civilian workers. Since the ACS never asks if a person is in the U.S. legally, you would have to make some inferences about the people whose records show not a citizen for the CIT variable. (There are researchers who have developed ways to estimate the number of undocumented workers, but I don't know how that is done - Google about. ) I would not recommend using a value of 7 for the COW variable to find gig-economy workers since people with small businesses are also covered by that COW value. So, I don't really know any way to isolate gig-economy workers in the ACS. Perhaps others can chime in with their thoughts on this.

    Doug Hillmer
  • In reply to Doug Hillmer:

    Great - thanks Doug. Yeah, gig-economy workers seem to be particularly hard to pin-down in ACS/PUMS data.