Press "Enter" to skip to content

State Labor Supply Seems to Under-Count Available Indian Workers

I continue to try to nail down how well South Dakota includes the American Indian population in its unemployment data. Following my weekend calculation that ignoring unemployment on the reservation causes us to underreport statewide unemployment by ten percentage points, an informed reader insisted to me that the state does count jobs data from reservation counties.

The state Department of Labor does include reservation counties in its Labor Supply report:

South Dakota
Labor Supply
July 2014
South Dakota 51,755 Deuel 250 Lawrence 1,625
Rapid City MSA 7,935 Dewey 900 Lyman 320
Sioux Falls MSA 14,505 Douglas 150 McPherson 130
Aurora 140 Edmunds 255 Marshall 215
Beadle 975 Fall River 505 Mellette 175
Bennett 200 Faulk 110 Miner 115
Bon Homme 255 Grant 430 Moody 430
Brookings 2,070 Gregory 205 Perkins 145
Brown 2,090 Haakon 115 Potter 125
Brule 305 Hamlin 380 Roberts 535
Buffalo 295 Hand 170 Sanborn 100
Butte 440 Hanson 245 Shannon 1,295
Campbell 80 Harding 90 Spink 355
Charles Mix 445 Hughes 1,105 Stanley 310
Clark 165 Hutchinson 300 Sully 105
Clay 890 Hyde 85 Todd 665
Codington 1,580 Jackson 255 Tripp 260
Corson 320 Jerauld 205 Union 1,035
Custer 805 Jones 105 Walworth 370
Davison 1,230 Kingsbury 275 Yankton 1,345
Day 335 Lake 615 Ziebach 215

Source: Labor Market Information Center, South Dakota Department of Labor and Regulation.

Now we're really going to apple-orange the numbers here. "Labor supply" is the state's guesstimate of the number of workers a new business setting up shop could get to apply for work in a given county. It includes workers who would leave their current jobs, folks who would commute from elsewhere, and folks who aren't working but would like the chance to work. My Saturday calculations came from tribal reports that included folks who are "available to work," whether they are actively seeking jobs or not. Those tribal numbers do not appear to count folks with jobs and willing commuters from elsewhere. So the state's "labor supply" numbers should be larger than the figures the tribes report.

Furthermore, my numbers are based on reservations; the state's numbers are based on counties. Some reservations partially touch multiple counties. If we try to line up reservation numbers and county numbers, some of those counts (i.e., on the Crow Creek, Lower Brule, and Sisseton-Wahpeton reservations) will include some non-Indian labor supply. So again, the state's numbers should be larger than the tribes' numbers.

But make spreadsheets boldly, said Luther. So here we go!

Reservation County State Labor Supply Tribal Unemployment
Pine Ridge Shannon 1,295
Pine Ridge Jackson 255 26,408
Standing Rock Corson 320 3,074
Cheyenne River Dewey 900
Cheyenne River Ziebach 215 9,893
Rosebud Todd 665 11,909
Yankton Charles Mix 445 88
Lower Brule Lyman 320
Lower Brule Stanley 310 759
Crow Creek Hughes 1,105
Crow Creek Hyde 85
Crow Creek Buffalo 295 380
Sisseton-Wahpeton Marshall 215
Sisseton-Wahpeton Day 335
Sisseton-Wahpeton Roberts 535 6,023
Flandreau Moody 430 472
total 6,430 59,006

On the reservations, my tribal-report-based data shows 59,006 people available for work. In the state's broader calculation of labor supply, in counties either entirely or partially located on reservation land, including some people who already have jobs and who would be willing to drive to those places from elsewhere to work, we find 6,430 people available for new jobs in July 2014. Statewide, the official labor supply is 51,755.

Something is missing here. Either the tribal reports of folks available for work are way off, or South Dakota really is excluding Indians from its labor statistics.


  1. Wayne B. 2014.09.08

    Cory, do you know how the tribal report calculates its unemployment?

    Note the LMIC generates its Labor Supply through the "handbook" method, rather than actual surveys.

    If we review the technical notes:

    Labor Supply = Labor Supply of Underemployed + Labor Supply Not Employed

    And Labor Supply Not Employed = Unemployed + Estimate of those currently not in the labor force, but have searched for work in the past year and are available for work.

    Both qualifications require the person to be relatively actively looking for work. It's wholly possible we have a mismatch of definitions - where the tribal report uses unemployment the way you and I may think of it - people without jobs, yet the LMIC uses the federal consideration for unemployment where you have to be looking for a job.

    Question: what does federal data say about those counties' unemployment rates? Does it jive?

  2. caheidelberger Post author | 2014.09.08

    You may be so bold! Thank you, and my apologies to folks on both reservations for flipping your counties. I have changed the chart to align counties and reservations properly.

  3. caheidelberger Post author | 2014.09.08

    Very possible, Wayne. I'm not clear on the methodology the tribes used for the data they submitted to BIA. More reading!

  4. Wayne B. 2014.09.08

    This may have something to do with that big long caveat the lone economist provides about not using tribal unemployment data as an apple-to-apple comparison point...

  5. Bob Klein 2014.09.08

    Even in Todd County, # of people looking for work does not necessarily equal # of tribal members looking for work. There are both non tribal members and members of other tribes resident in Todd County.

Comments are closed.