Wage Records and Data Sharing Regional Meeting
Minutes Draft #1
Casper College: Skip Gillum, Vice President, Academic Affairs
Colorado Dept. of Labor and Employment: Ben Garcia, Web/Database Administrator and Clay Bundy, Research Analyst
Iowa Workforce Development: Pat Callan, Actuary, Workforce Research Bureau
Montana Dept. of Labor and Industry: Bob Liffring, Economist/Statistician
Nebraska Workforce Development-LMI: Phil Baker, Administrator; Mary Findlay, Research Analyst; and Duncan Hsu, Research Analyst
New Mexico Dept. of Labor: Dan Hall, Economist Supervisor
North Dakota Employment Statistics and Workforce Programs: Warren Boyd, Information Services Supervisor, and Tom Morth, Research Analyst
South Dakota Dept. of Labor, LMI Center: Phil George, Director; Laura Sickmeller, Research Analyst; and, Loren Harms, Statistical Program Manager
Utah Dept. of Workforce Services: John Mathews, Economist
Wyoming Dept. of Employment: Wendy Tyson, Administrator, Employment Tax Division; Tom Gallagher, Manager, Research & Planning; Alfrieda Gonzales, Administrator, Workforce Services; Mary Peterson, UI Employer Services Supervisor; Randy Hopper, Employment Program Supervisor; Tony Glover, Economist; Sylvia Jones, Economist; Sara Saulcy, Economist; and Susan Murray, Administrative Specialist.
WAGE RECORDS AND DATA SHARING REGIONAL MEETING
August 1, 2001
Track 1: Getting Started: Managing and Organizing Administrative Data (Need to Have
MOU with Wyoming to Participate)
Meeting Place: 246 South Center, 2nd
Contact Person: Craig Henderson (307) 473-3822
Susan Murray (307) 473-3807
Focus Topics: Obtaining drivers license data,
Title I data, vocational rehabilitation data, other
Load Mew Mexico's (and any other state) 2000 First Quarter Wage Records on
to the Wyoming Server, matching against Wyoming wage records, running
Editing the wage records, maintaining text files, frequency of
quarterly meetings with UI Tax staff
Variables created for research purposes, i.e. primary employer, number of
employers, utilization index
Track 2 On the Road: Using Administrative Data To Meet User Needs
Meeting Place: Employment
Security Commission Building
100 West Midwest, 1st Floor Conference Room
Contact Person: Tom Gallagher (307) 473-3801
Susan Murray (307) 473-3807
Focus Topics: Meeting workforce
development needs, Using administrative data, integrating
wage records with other events, and sharing data.
8:30 AM Welcome, Opening Comments and Introductions Tom Gallagher
9:00 AM Report on Track 1 Activities
9:15 AM Purpose and Goals Of Regional Meeting Tom Gallagher
10:00 AM Break
WIA Consumer Reports
Casper College Pilot: Meeting WIA Section 122d Requirements Tom Gallagher
Employer Survey Results Sara Saulcy
WIA Consumer Report Sara Saulcy
The Community College Perspective
Skip Gillum, PhD Vice-President Casper College
11:45 AM Response from States Phil George SD Follow-up Project
12:00 Noon Lunch on your own
August 1 Track 2 (continued)
1:15 PM State Roundtable-Where are we at and where do we want to be?
Wage Record Quality Control
Wyoming’s Quarterly Download Meetings Craig Henderson
3:00 PM Break
3:15 PM Report on Tract 1 Activities
Sharing of Wage Records: Common Standards & Edits MOUs, etc.
Facilitator-Phil George & Tom Gallagher
August 2, 2001
Track 2 On the Road: Using Administrative Data To Meet User Needs
Place: University of
Wyoming School of Extended Studies & Public Service
951 N. Poplar, (307) 261-2250
Contact Person: Tom Gallagher (307) 473-3801
Susan Murray (307) 473-3807
Wyoming Turnover and Instability Index Tony Glover
Growing Oil & Gas Industry: Where do the Workers Come From
Using Administrative Data to Anticipate Labor Shortages
Using Drivers License Data to Complement Wage Records9:15 AM Using
Drivers License Data to Complement Wage Records
Using Drivers License Data to Complement Wage Records
Nebraska, South Dakota, Wyoming staff
10:00 AM Break
Administrative Data Uses Today and in the Next Decade
State Workforce Development Perspective
Alfrieda Gonzales Wyoming Workforce Development Council
11:45 AM Questions and Answers
12:30 PM Lunch
2:00 PM On the Road: Where do wego from here, a review of the NASWA LMI
August 1, 2001
I. Introductions and overview of status of participating states
II. Purpose and Goals of Regional Meeting
Two years ago Region 7 and 8 states met and one outcome was separate Benefits Surveys published by Nebraska, South Dakota and Wyoming. Goal was to publish the survey following one format, although we did not achieve that goal, at least we were able to come up with separate surveys.
Phil Baker mentioned that New Hampshire pushed forward a proposal for developing a consortium of states doing a benefits survey, and ETA approved funding the work group. A study will be done soon on what common things we want to do across all states. ETA & BLS will be involved in the process. BLS wrote a proposal and presented it to the WIC. They will research what's out there, what works, and what doesn't work, and will develop some prototypes to be tested by some states.
Wyoming’s strategy is to use IRS forms (because we know they have the information in the drawer); we know where they collected the [Benefits Survey] data and how they have it. We documented our strategy in our publication and on our Internet site.
A discussion draft of "Measuring the Success of Education, Training, and Workforce Development on Internet", by the National Association of Workforce Agencies, Labor Market Information Committee, is on the Wyoming Internet site at:
II. Purpose and Goals of Regional Meeting
In the July 2001 Wyoming Labor Force Trends, we have documented our pilot WIA Consumer Report with Casper College [note: the copies passed out were near final drafts of the publication which should be published shortly]. We relied on what SD learned when doing follow-up with students.
Wyoming also passed out Wyoming Wage Records 1992-1998 and Tracking Job Changers in Wyoming Economy. When WIA was coming in, we figured it was time, in terms of accountability, to show what we had done. Job Changers was a cohort analysis over time. When tracking a group (students, TANF completers, etc.), the same issues are going to crop up repeatedly.
Wyoming also handed out Outlook 2000: Detailed Occupational Projections and Labor Supply. Documents our attempt to use administrative data to address the gap in the process of measuring the level of utilization of current people in the workforce in order to address Workforce Development Council’s questions.
Discussion about defining turnover, churning, labor dynamics. These are new concepts for which we do not have the language. We are being forced to abandon a lot of the old concepts we have, and to explain the new concepts. Phil G. would be interested in a state doing something with vacancies (teachers, nurses, etc.) It is said that vacancies for these positions are high; high relative to what? Phil G. would be interested in seeing how job vacancies compare to churning.
One question Wyoming Department of Education and the Governor have is are the teachers being paid enough to retain them - to prevent turnover and out-migration. We saw an enormous increase on the first quarter for number of hires and exits (handout from Tom G: Table 2: SIC 8211 Own 30 All UI Accounts), due probably to the school district using substitutes more often in that quarter. A large number of teachers are off during the collection period for coaching. So, during the OES collection they include payroll for substitutes as well as teachers, thereby pulling the wages down in the OES collection.
III. WIA Consumer Reports
Introduction to WIA consumer reports progress in Wyoming by Tom G. Fortunately we had a good relationship with Casper College and had the funding to take them on as a client. But, we had to learn about them. Craig Henderson and Sara Saulcy spent some time with their staff. Became a model to use when dealing with other colleges. Hope that the previously mentioned Trends article will point out to other providers that there is a benefit to working together. Our community colleges are almost autonomous rather than being solely a state entity as in other states. We’ve done MOU’s with three colleges now, and a fourth e-mailed us that they are interested.
There is a section in FERPA allowing the colleges to participate in confidential SSN sharing in order to benefit and improve training programs. WIA set up a system where you could use Wage Records ("WR") to make manifest how training is used in the market. Then you can look at improving programs.
Sara Saulcy provided overview of WIA Section 122d and discussed the Trends articles about the Consumer Reports. In the pilot, we only did follow up on graduates. Colleges have the choice of providing the information on outcomes themselves, but may have limited resources to complete those tasks. We piloted the survey to see how it works in practice. We used time frame of six months after graduation rather than different times that they get the jobs (six months after each person gets a job). Aggregations (i.e. CIP and OES codes) in Figure 1, P.12 are somewhat arbitrary - tried to lump together with like categories, but some aren’t (e.g. food prep and protective services) due to confidentiality restrictions.
Next time, we will be bringing in students who didn’t graduate but received some training (not just completers). Some of the students may also have gone onto another institute of higher education. Section 122d specifies follow-up of program participants, not just "completers". Next survey should be mailed in early January. As we expand there will be greater interest in doing longer-term follow up.
Tom G. elaborated that some of the employers did not want to give information on satisfaction with an employee per their personnel policy. This may also occur with other WIA surveys. We’ve learned that some things required by WIA are just not practical.
IV. The Community College Perspective, Skip Gillum, PhD, Vice President Academic Affairs, Casper College
Skip & Tom went to a meeting a year ago where many college Presidents were upset. The Presidents said they were doing illegal things (Family Educational Rights and Privacy Act of 1974, "FERPA") trying to get this information.
Skip said there were two possible consequences if Casper College was in violation of FERPA:
1) The school could lose federal funding. From research, Skip found that very seldom did entities lose their funding. Some presidents asked their counsels and were told they would be in violation. Skip questioned his council and the attorney general and they said they could do this research because it is in the benefit of the students to improve education training programs.
2) Could get reprimanded. Easier to ask forgiveness than to wait for the process to begin. Casper College decided to implement the Consumer Survey, apologize if they had to, and live with it the consequences.
Neither of these consequences have occurred. Before, they had to rely upon information from the student, and information may have been suspect (may have been embarrassed to say they weren’t working, weren’t making much, etc). If they got a 10 percent survey response rate, they were doing well. Student may have been wary about being completely honest of their opinions rather than if third party surveyed them.
In Wyoming there was a rumor that the state was suffering from brain drain, that there was out-migration and almost no one was staying in Wyoming. Casper College had found that 70% of their students were finding work in Wyoming. They were making as much or more than their counterpart who hadn’t gone to college. They broke down by age and gender compared to other counterparts in community. Comparatively, women were making more (45.6) than the other women who hadn’t gone to college. [Wyoming Department of Employment, Research & Planning, A Consumer’s Guide to Educational Outcomes, draft-May 22, 2001.]
Benefits to this type of survey was that he could call Research and Planning (R&P) and ask them to try finding more data (i.e. a quarter before, year before, or two years before they were in the Casper College file). What is the value of someone coming to college? Is there a payback? This allowed us to have a better look at that because we had a look at individuals before college, during, and after. Wouldn’t mind tracking this 1997 cohort in a couple of years to see where they’re at then.
Tom G. noted that the Consumer Report Appendixes give a lot of the information that Skip is talking about.
Skip stated that some of the problems we had to be conscious about were that their graduates who went to the University of Wyoming ("U.W.") may be in the job market because they’re working at a minimum wage job while furthering their education. However, UW is not cooperating, so the wages of those students may pull down the Casper College averages. Skip thinks Casper College may be able to get around that problem by tracking those graduates who’ve had transcripts sent to U.W. – this would work unless there are multiple people with one name (e.g. three Mary Smith’s). Tom G. added that if a Research Office doesn’t have all of the information from higher education, we are going to misrepresent and send out false information. We would like to report on the persons going to U.W., but report them separately. Skip also pointed out that the pie charts of where Wyoming works vs. where do graduates of Casper College work matched up extremely well. When Skip first read the wage data of completers, he panicked. However, when he read the report, he saw that it was in par with the rest of the state. Tom G. stressed that if we give our customers an analysis of how the market works and how each institution or program functions within the market, it is less threatening than just handing out data and not explaining it.
Skip hypothesized that the possibilities of interacting with the state with this type of data are endless. Questions people have to ask are at what point in time does a student maximize his education? At which point in time have they developed the skills that they need so they can go out and get employed? Casper College couldn’t get anyone to stay in the medical transcription program until graduation because employers were grabbing them once they had a certain amount of education/hours/training. Do they do they get a better job at 15 credits, 30, 60, etc. Do they interact better if they’re straight A, B, C student? Community colleges in this state are being tapped more and more to train for industry that is coming in. The employer can qualify for workforce training funds and can choose to use at the institution they want to use the funds at.
Colleges are supposed to prepare people to go to work. If they’re training them to go to work, did they succeed? Did they make more money than if they didn’t go to school? Need to see if they impacted the students. Other colleges would tell you they are supposed to make students better educated and better people. My goal is to make an impact on their salary or we didn’t do our job.
We haven’t yet presented this study to the legislature, but Skip is working on it. He is first presenting it next week to the local service clubs. He knows college impacts if the student gets, e.g., a nursing degree, but now he’ll be able to show that other fields make more money as well.
Tom G. Figure 4 in the Casper College Consumer Report is a comparison with similar cohorts; same age, gender, and earnings history. The earnings differential doesn’t begin to emerge until three to six quarters after graduation. This gap suggests, from a workforce development viewpoint, that if you could reduce the number of quarters from where it may be that graduates are conducting a job search, that would be a good improvement.
V. South Dakota Follow-up Project
Phil G.: the genesis of the South Dakota follow-up project was in 1988; a group got together and figured LMI was right agency to follow-up on graduates of several programs (JTPA, public higher-ed, voc-rehab, etc.). LMI gets graduate and enrollment data from all of the programs, then takes the information and tries to figure out what’s happened. At first, the programs wanted to wait a year to conduct follow-up, but they didn’t realize the nearly two year time lag involved in follow-up after graduation date. First the LMI shop gets SSN’s and runs them against the WR file. They find quite a few in SD, especially voc and tech individuals. They do a survey similar to Wyoming’s, but SD asks for a hire date (in addition to starting wage, current wage, occupational job title, place of work, and benefits), but they keep the form to a maximum one-page length.
Some employers get surveys for just one graduate, others may get them for a hundred graduates. Employers don’t know the genesis of that person (GED, voc-rehab, etc). South Dakota has been surveying for 10,000 to 11,000 graduates per year. Some people who wanted to wait a year before surveying decided the time lag for the data was too long and decided to survey after six months instead. Because they shortened the time period, wages dropped dramatically in most cases - similar to what Wyoming found for Casper College graduates. They shoot for a 90% response rate, and ended up with about 83% last year - normally it is 85% or higher. The first couple of weeks the response was incredible.
Now SD is now getting SSN’s for secondary vocational completers and high school completers to conduct follow-up. A big issue in SD is the opinion many have that HS completers aren’t ready for workforce. Some of the data will be helpful to understand how many get jobs, how many are enrolled in school, and so on. They cannot require students to provide an SSN; they get them through the Department of Education (who obtain them through the local school districts). SD ended up with about 80 percent of the SSN’s, though. The education people wanted the SD LMI office to do the study of H.S. completers. Follow-up of H.S. students is done strictly with wage record, and aren’t included in an employer survey. Phil G. was told that they couldn’t do the study without parental permission; it is up to the school district to get the permission, but about 80 percent are getting SSN’s with permission.
One thing that SD does differently than Wyoming is that after they do comparison to their WR and other states, they generate a "missing graduates" file. They provide that list back to the placement people. We have good working relationship with technical institutes. Higher ed is more complicated - hard to sit down with everyone. Accounting for about 90 percent. This activity is being funded by the agencies and entities that are interested in this information. The fee structure is based on a base cost plus cost of records. They have two compartments: the complete follow-up and wage records. This began because of Governor Jenko’s requirement ("the governor says"). The technical institutes could not begin a new program unless this survey showed a certain success rate.
Another thing they want to look at is starting wage vs. wage they get about a year later.
They only survey graduates, not longitudinally (e.g. follow-up of 1997 graduates at a later time).
South Dakota has also been getting a lot of ad-hoc requests. Also, placement staff with whom they’ve interacted with have additional requests. A lot of the higher-ed institutions have placement staff that have been getting high response rate for many years (through placement interviews at graduation) - it takes a long time to build relationship with those entities.
Skip said educators have long just felt (without data to back it up) that there was a value to education. But, they can’t improve unless research like this is done.
Loren handed out report on a combination of the four technical institutes, South Dakota Follow-Up Project.
For the category "Graduates Residing In South Dakota", they ran the graduate information against SD driver’s license file. Tom G. asked how they’re handling issues of analysis. Phil G. elaborated that they need to add narrative with the reports, but they’ve been busy and haven’t hired somebody to do the narrative. They meet with the program representatives to discuss the data and what it means.
SD gets the job titles from OPM, DOD, etc.; in most cases they have to have an occupational job title. When they match the person in another state, they don’t have a job title, but they still count them as with jobs.
The process of deciding whether an occupation is related to training is still being developed, but basically, with technical institution staff they build a table based on relationship. However, there are cases that doesn’t work good (i.e. a drafter employed by Maynards coded as Sales Associate by the store, but the drafter uses the training received). There are always exceptions. They have some cases of higher-ed wanting to count an RN degree in an LPN job, but other higher-eds do not want to because LPN is not up to RN. But, they are trying to work on some consistency.
VI. State Roundtable - Where are we at and where do we want to be?
Wyoming (by Tom G.): Overhead from Tom G. - Framework for the Analysis of Labor Dynamics w/Administrative Records (see attachment to the minutes). When looking at this table, should start at the lower right hand side "Start Here". Developing concepts and ways we manage data is where most work needs to be done. One of the interactions listed on the table is the Workforce Development System.
We have the capacity to understand which individuals obtained which jobs in a new firm (i.e. Lowe’s in Cheyenne). We could have the capacity to do microanalysis of what happens when a firm comes in/opens. As one part of the economy grows, where does the staffing come from? How does your economy function - next question is what is the role of each of these pieces? Takes us down towards lower right-hand side of [Framework] table.
Tony Glover, "The Flow of Labor in Wyoming", Wyoming Labor Force Trends, March 2000.
After years of trying, we finally got JTPA data. Need to draw control groups that have some reasonable comparison qualities to the group you’re studying. This relates to the Framework table; the real challenges are in the middle two columns of the [Framework] table.
North Dakota (by Tom M.): They operate as a State Occupational Information Coordinating council primarily under Job Service, and they date back to 1994. They have an MOU with SD, and are doing an employer survey somewhat similar to SD. They have access to ACT tests - approximate cohort to represent an entire year’s graduates. This is a good approximation to determine placement. They receive the files from Job Service, and enrollment files from their higher-ed department. ND is the only state that SD has an enrollment MOU with. ND is looking at securing institutional information (i.e. tribal). Their state income department also provides an aggregate report/final follow-up when have exhausted all other avenues. Pointed out that it doesn’t necessarily mean they’re a ND resident if they file ND income tax, though. ND does not have access to the state drivers license file; they feel income tax file is enough. Also, they are getting demographic information from the schools. Their focus is on groups of students; they do not conduct analysis on all wage records in the state. If the customer want an analysis by categorical data, they have to provide the data to them.
North Dakota (by Warren Boyd):
Same as what Tom Morth explained - it is the same in their LMI shop.
New Mexico (by Dan Hall): They are just getting started and are small.
Montana (by Bob Liffring): Montana doesn’t have an MOU with anyone right now, but they do have a draft they’ll take to their attorneys. They date back to 1993. They have no demographic information with records, but are investigating use of DMV records. People are shying away from using SSN’s on DL’s though, so now the DMV will be assigning an arbitrary number to each person instead of using the SSN. They’ve linked wage records to ES-202, and have done some internal research on a micro basis. They’re responded to some requests for some special research projects. They have been providing WR information ("dummied-up SSN’s") with Washington state, but he doesn’t know if they have an MOU. They will have an MOU before providing any further information.
Colorado (by Ben Garcia):
A colleague answers a lot of ad-hoc requests. They are working on MOU with their community college system, and they have a feature built into the their eligibility website where trainers can enter SSN’s for their WIA performance information. They have done some analysis of WR.
Nebraska (by Mary Findley):
They are starting with wage records. They have a work group with Labor and Post-Secondary, and have conducted a test run with three colleges. Working on what reports, maps, etc. to get back to them. At this time, they are just getting completers, not enrollment. The community colleges will classify those programs they wouldn’t expect people to go to work in. Nebraska also has agreements with a few of the surrounding states. There will be some information forthcoming to Wyoming by the end of August. They haven’t had a chance to match with 2001 1Q EQUI. They have some follow-up of teacher graduates to find out if they’re working in any of the other states, and if they’re in the teaching field, or what field they are in. Have WR for 1996-present.
Nebraska (continued by Phil B.):
Nebraska has recently worked with the Department. of Motor Vehicles; they don’t use SSN, although they do collect them. They match the DMV data, which gives demographics with wage records. They are just getting started. NB did the traditional looking at male/female wages and vocations. There is a large demographic difference in wages between, i.e., Scottsbluff and Omaha. Looking at wages along the interstate, they get increasingly larger as closer to Omaha. It is obvious from the shading on their map that their graduates are gravitating towards Iowa. They are going to release this data later this month. They are also able to throw in age categories. FIRE ages 50+ was the highest wage category for Nebraska.
South Dakota (by Phil G.): They are well along with their follow-up project, and now need to integrate need for WR and performance measures. With training providers, they need to work into how they collect performance information and WR . They are working with Colorado to use their Navigator application, which a private vendor modified for them. This is an avenue to input their consumer reports in. The next step for their office is going to private universities and tribal government, but there is an issue of how to fund inclusion of that data. Their goal is moving from public to including private entities.
Utah (by John M.):
Utah hasn’t given data to Wyoming yet. The last three to four years, because of accountability, their Department of Education has asked them to take on the verification responsibility that the Department of Education used to be in charge of. They provide the standard format of supply for higher-ed (not high school), which creates an output file that has a lot of data relative to WR. Since they don’t have hire data like they used to (haven’t since mid-90's), they send an exit date for their graduates, and have them verify placement & retention. They run the data in January for June, then run again in April. All he does is send them a database; reports may come later with WIA reporting. The next step will be to provide private providers data. They will probably charge up to custom designed reports, but they are still dealing with some basic definitions. They are trying to add two fields each year to Wage Records. Changing the WR means changing UI laws in state, though.
Iowa (by Pat Callan):
Primarily they have a customer-tracking agenda. Their department includes WIA, Workers Comp, etc. Their office used DMV records to create a gender equity survey. They didn’t receive any money for that, but they published the report (it may be on their website). From a customer-tracking standpoint, legislature and partners are still nervous about having "outsiders" analyze results. They have to provide SSN’s and they get WR back, and there is a small fee involved in running jobs. Their office has to have contract with people i.e. the Department of Human Services. Their legal council wants everyone who gets the data to sign it. They’ve done general research on patterns of wages. UI are the ones who own the data. They don’t collect taxes or benefits on the information, so it’s not programmed in to obtain that information. They get some information on DMV, but there are some limits between wage files and DMV files. They don’t have demographic info - where the people work. Pat doesn’t know where they are with sharing with other states. They share with Illinois, but do not mass-share. No one has mentioned WRIS, and Pat doesn’t know their policy on it. They have an outside agency do data entry, so errors are not found and data doesn’t get corrected unless a person files a claim.
VII. Wage Record Quality Control and Wyoming’s Quarterly Download Meetings
Craig H. passed out a handout titled "UI Account Appearance WR vs. EQUI and 1999 Q1 Wage Records."
In March 2000, R&P decided to formalize their presentation of the information, and began using a summary report as a way to have some "talking" points. People in R&P, 202 refiling, multiple worksite function, etc. participate in a quarterly download meeting.
The first and third pages of the handout represent quarters, and the second page is 1999 annual information. We look for duplicate SSN’s that have been reported, and when a business changes ownership and both report the same wages for the individuals, we get rid of predecessor. Also, if different wages for the two, we address those issues. We are trying to show the differences between the two databases, and are trying to bring these to the attention of the Field Tax section. If we don’t catch problems in first quarter, may be able to catch before final quarter. We are also trying to improve intake of data and follow-up with the data. First quarter of 1999 (figure) represents resolving problems with the new imaging system.
Wendy and Mary (UI Tax) addressed a question about quality control of the data entry. We have had double entry. In addition, we’ve brought in the data processing staff, benefits staff who work with special claims, etc. UI Tax is using these meetings as a training tool to point out the importance of this data. One thing that happened was a computer glitch that added $230,000 in wages for each employee. A cooperative effort with R&P identified how big the problem, how widespread it was. It is supposed to balance to a certain figure, but things can happen after. Wendy likes Pat Callan’s idea of building in checks in the computer system that will point out possible errors. We’re trying to get zip codes. We classify a task code, coverage type, new hire date, and hour column which is used by Workers Comp. The hours section is probably the messiest thing. We have a lot of half hours, and figures that are not rounded. Employers don’t report consistently, they report in different manners.
To help with the editing, Tom G. pointed out, Tony set up a process to help with the editing of downloading WR earnings over the lifetime of the individual. Earnings outside the average are flagged. Tom said that at least we have a structure in place where we can address the editing process.
Wendy thinks the process has worked well all-around. We’ve knocked off some ugly non-respondents. Any WR program has to have a major edit.
VIII. Sharing of Wage Records: Common Standards & Edits, MOU’s, etc.
Tom G. distributed copies of a draft Colorado/Wyoming MOU.
Phil stated that one of the things SD wants to use the agreements for is Wage Records. The following are items Phil listed that may be standard fields that would be useful with WR:
Job Placement and Enrollment Data Format
Last Name Alpha 30 characters
First Name Alpha 30 characters
Middle Name Alpha 30 characters
Year Alpha 4 characters
Quarter Alpha 1 characters
UI account # Alpha 15 characters
Industry (NAICS) Alpha 6 characters
Ownership Alpha 1 character
State ID (FIPS) Alpha 2 characters
County ID (FIPS) Alpha 3 characters
Resident of state Yes or No 3 characters
Enrolled in State Yes or No 3 characters
File should be fixed-width ASCII text file.
If data is not available or redisclosure of confidential data, then blank filled. Include multiple wages records if have multiple employers for quarter.
Tom G.: Colorado/Wyoming MOU Colorado, page 3 of 7, D.2)d. - may or may not be available - a state may not be managing this. If you move from a two year college to a college in Colorado, this is something we’d like to track.
Tom G. also mentioned that all of Wyoming’s agreements prohibit redisclosure (if we get the data we can’t redisclose it). The information can only be released in statistical (grouped) form, so individuals cannot be deduced from the data. Page b-1 of the Wage Records publication lists permissible uses of the information a state collects on a driver’s license file. Cannot contact the individual directly - there are penalties for this. That’s why we publish the documentation - to make available information about what we can and cannot do. There is a recommendation regarding suppression in Chapter Six of the draft "Measuring the Success of Education, Training, and Workforce Development on Internet" that is now on our website (see page 4 of the minutes). We do want the name if we can get it for accuracy purposes
Bob Liffring stressed that it should always be in ASCII text - a specific type of file.
Tom G. pointed out that the frequency of exchange should be in the MOU.
August 2, 2001
IX. Labor Turnover; Wyoming Turnover and Instability Index by Tony Glover
a) Turnover (handout by Tony G. "Turnover, Instability Index, Oil &
A group of people had started this before he began w/R&P. Tony queried the current quarter, quarter before and quarter after to see where they were. He expanded to include a new hire and re-hire categories. If they worked sometime within the period, classified as rehire. If they hadn’t worked sometime within the period, they were classified as a new hire. Page two is a timeline of an individual’s interaction with an employer. "empW" is a proxy for UI account. Employer one is the primary employer (by highest wages) and so on. The second panel is the explanation of the graphic above it. The UI account number changes completely if someone buys a store, just change last 3 digits of UI account when Tony maintains the UI account number, but creates an employer number with last 3 digits of the UI account. Pat C. mentioned that when they [Iowa] assign a new number, they assign a complete new number. SD also assigns a complete new number - that is the biggest issue they have. Tony plans to permanently link the predecessor/successor in time. We are starting to address the issues. In Wyoming’s 202, when the whole UI account is changed, they don’t always seem to fit very good. If a local restaurant is sold to another restaurant, tracks if fifteen people are all working for one employer one quarter and the same fifteen are working for another employer in the next quarter.
Tom G. - back to issue of where are the common problems; if there are common problems, maybe we need to address them. He doesn’t find definition of turnover addressed effectively. Tony pulls out births and death from turnover computation. Phil G: there may be an issue at BLS because these changes in businesses may not be economic but administrative (different owner). Taking the firm as the unit of analysis is not something we’ve done at all well until we get to the instability index.
Tony stressed that the flow rate will vary with seasonal industries. Continuous rate is sort of a level of stability with the employers in that industry. Tom G. said whether or not we use 202 level or sum of SSN’s, 202 allows you to link back to the time series. You don’t have to explain to other people as much as if you use the sum of SSN’s. Could also do table on p. 37 of Outlook 2000 by county level People in economic development really want this, but if they really want county turnover, they’ll have to pay for it. There is a good reason to use sum of ssn’s-human resource costs. How much did it cost me to hire these 9 people or these 9,000? Bob L. pointed out that there are vast differences between using WR to estimate total employment vs. 202. Tom G. said that’s one of the edits - everything we end up doing ends up being an edit. Phil G. said that SD uses use total WR, and to account for multiple worksites, they have a formula that allocates based on 202 numbers. SD looks at change in employment numbers for a county; it’s not perfect but one way they found of doing it. Use change of employment to allocate to a worksite. Tony would like to compare turnover rate using 202 vs. employment vs. residence/location. Tony said that turnover can be looked at on an industry level thing - it is not unstable to change from one firm in an industry to another.
b) Instability Index of Individual
Does the training program or government program give some sort of stability to the individual? Looked at how to measure unstable or stable activity. Stable activity = continuous employment. Taking any point in activity that was an entry or exit. Any entry or exit is seen as unstability. Page six summarized that activity, during that fourth quarter that person had 2 hires, 3 exits, and so on. Tony’s March 2000 Trends article applied to the Department of Family Services, Division of Vocational Rehabilitation and JTPA clients.
Page 7 shows that you can aggregate by employer. DVR has a six month follow-up. Entries=hires. This is all SSN’s for all the employers, not just completers.
Phil G. asked why Tony categorizes an entry and exit in one quarter as something else?
Tony explained that you can pull out people, i.e. in Jackson area that work for two months and drop out. When he looked at it, he could see that a lot of those workers were from Albany County (where the University of Wyoming is located) - could drop those out.
Ben G. asked if we would work with a Workforce Development Board on this, and Tom G. said that we need to be explaining to them the instability index, and do some modeling between what we’ve collected in wage index and benefits survey. "These are characteristics of the firms that get and retain employees." This firm in this SIC that is this size that pays this average wage fits this profile.
Page 8 shows how the file exists when we download it; when matching the 202 and DL data, only match about 84% of the data. An individual who has a Cheyenne Driver’s license may work in Natrona County (i.e. Tony).
X. Growing Oil & Gas Industry: Where do the Workers Come From? Tony Glover
Tom G. elaborated that Wyoming has had very little growth in the past ten years; we even had one point in the last ten years that he would call a recession. Recently, the energy prices went up, and in the boom/bust cycle, we’re in beginning of the rat stage. We have one of lowest unemployment rates, but where are we going to get the labor to fuel this growth? If this boom goes any further, wages will continue being bid up and people will move from other industries to those jobs. This leaves the Workforce Development question on where are the vacancies?
Coalbed methane is the boom this time. Very shallow water wells, compressor stations being run by on-site diesels, collection without infastructure. Traditional oil and gas exploration employ higher-tech. A good deal of the job orders/OES are for low-tech bodies going out to drill water wells. Per Tony, a lot of the wells are being tapped for anticipation purposes.
We can aggregate on different levels; one level he aggregates on is industry - choose primary industry. Pulled people who entered oil/gas as primary industry, (this is very preliminary) (p11), plus they had to maintain that employment for more than three quarters (this table doesn’t reflect that, though), can then group by age & sex. This is the first time they have ever worked in the Oil and Gas industry. Many of the people entering the field are younger males coming from retail trade.
Tom pointed out that the destination of this table could by county instead of industry. This table includes time but not space. In terms of the form of analysis, this has application to many problems. This could be restricted to a county regardless of industry. Where is the employment coming from that is stocking an MSA? Also, geographic difference in earnings; Omaha vs. satellite areas. Surveying to find out what wage they would move for would be mute because you can say what they did move for from the wage records. One thing that happens in WR is the retention period gets longer. Who knows how to analyze CES without this kind of data behind it.
XI. Statewide Commuting Patterns, by Tony Glover
From Driver’s License, he knew where the residences were. The unknown category ended up being the largest (p. 12 of handout) - compared 202 data. We have been discussing an MOU with Idaho. One question is whether or not people are enrolled , but another question would be whether or not there is a DL hit. One problem with the DL files are that you only have to renew your license x number of years (4 in Wyoming). Bob L. said he would be uncomfortable sharing DL data with us, and Tom G. said we wouldn’t want to know individual information, we just want to know if the other state has some reason to believe they’re a resident of that state.
Wyoming is going to publish the commuting article about Jackson next month. We have some people who have worked in Gillette for 44 quarters but never had a Wyoming DL, nor a UI claim. We’re going to assume that they are residents of Montana. Bob L. said Montana has a situation where mines have closed, people work and live in the center of state for seven days, return home for three, then go back to work for seven. What is their place of residence? That is similar to what is happening in the Gillette area; some type of housing is set up for the workers.
XII. Using Administrative Data to Anticipate Labor Shortages (by Sylvia Jones)
Sylvia has started using wage records as a means to anticipate labor shortages (by way of occupation imputation). Tom G. asked Sylvia to use OES material and look at it differently; he wants to be able to somehow infer from wages what the worker does. Sylvia had to make a lot of assumptions; she opted to go down to hourly wage by estimating hours in the quarter, so there are problems with that. At least 500 people that made 4 cents an hour, and 83 were making over $300 per hour, so she eliminated the complete ends. She looked at just transportation, not all of TCPU. When figuring the hours per quarter, she calculated 173.333 hours per month x 3 for quarter, but cannot necessarily assume the workers are full-time employees, either. Tom G. added that the objective was to rank OES wage rates from low to high, take WR, kick out anyone not in the 4th quarter, then take OES wage rates, and correlate occupations reasonably tied to that level. Referring to p. 3, "SIC 46 Distribution of CPS Age Groups", Tom said if you randomly assign OES codes to WR, you have some way of identifying occupations that are moving to Oil and Gas. Every [Casper College] survey that we do asks the employers who hire college students what rate they hired them at. We know there may be exits coming out of these occupations going to Oil and Gas. We’ll be able to look at it and say whether the Oil and Gas employees are coming from Casper College, or if the companies are getting the employees from Colorado.
Sylvia said one would expect as a person gained more skills and aged, they would move into a managerial type job. However, the correlation was very small. She found no correlation between wage and age in this industry. Many questions need to be answered. Who from a demographic standpoint is entering and exiting this industry? Who makes up this industry? Who is it that finds a job in the first quarter? These tables are the market answer to what is happening in this industry. Training/workforce development question is how many people went through One-Stop? How many people came from Colorado? We should be developing the theory. Wants a way to analyze across time and across industries. This is new stuff.
The age groups were adopted from CPS, and we got the age data from our drivers license file.
Tom G. pointed out that the entrance and exits have the same earnings structure; the charts were done correctly because entrance and exits were less than the continuous, as it should be.
We’re going to try to develop a strategy to attach groups of occupations to WR over the next few months. The single biggest cost in WR is design strategy.
Tom G. pointed out that one problem with BLS is that nothing is designed from the beginning to complement another program. CES has hours, but only for production workers. In wage records we’ve got all firms, so the firm-desired behavior shouldn’t be there (i.e. firm like John Deere throwing off average wages). Since we never really analyze OES, we’ll find some interesting things when we apply ourselves to it. This all goes back to putting occupational descriptors on WR. If there is no way to put these on WR, that’s a reason to go the administrative records route. If we succeed, we’ve found a way to deal with Workforce Development in a way we hadn’t before.
XIII. Using Drivers License Data to Complement Wage Records, Nebraska, South Dakota, Wyoming, and Iowa
Phil G said South Dakota did six studies within a year and a half. They are addressing such questions as "Where do residents work" and "Where do workers come from?"
South Dakota started with the DMV file to find out about residents for a city/county location. They used Zip codes, and also use ES data & new hires data to help determine where workers live. There is a lot of manual work to it.
There was a question about using residency for the LAUS program? But, the problem is that not everyone has access to DMV detail.
South Dakota is getting zip code, DOB, and gender, but no names. They will get some names from telephone book.
John M. asked if the states at the WR meeting have trouble getting their state government worksite from the multiple worksite? Tom G. said we should be able to get zip code from payroll for place of work., but don’t take the site from the BLS multiple worksite. Instead take it a from licensed occupation perspective. Phil G. said that multiple job holders are high in their state and it benefits their LMI office when they’re trying to narrow down place of work for one of the jobs. Tony suggested that longitude/latitude picks the job closest to the employee’s residence as their place of work.
Bob asked if the census does commuting patterns. Phil G. replied that they do, but there is a huge time lag in getting the data. Things have changed significantly since 1990. Also, Tom G. added that commuting in April is not the same as commuting in November.
Phil G. said that using DMV data with WR will help if you think you have a problem with your LAUS program. It will also help with your economic developers.
Tom G. brought up that a problem with DMV/WR studies regards drivers license renewal time lag. (Drivers renew every four years in Wyoming, four to five years in Nebraska, and eight years in Montana.)
Bob L. commented that this is similar to market areas, and would be nice to use this like a weather map with percentages, distances, etc. Phil G. said the SD LMI office will eventually use this for labor availability information because people use GIS a lot better than the numbers. He then said that street addresses work better for SD because using zip codes gives one dot for 8,000 people, i.e.
On the last page of the SD handout, they compared their numbers with census numbers. The numbers compared real well - they were very close. The last table shows that SD’s raw numbers were based their estimates on 89,899, while the census bureau estimated on 15,314 (long forms).
Pat C. said marketing research has moved to using secondary data. The project entailing his handout, Laborshed Analysis, is primarily a phone survey which largely funded by local developers. The concept is asking employers for zip codes and doing analysis of where the workers come from. A problem they have is large areas. It is a lot of work for local development group to get zip codes from the employers in large areas (i.e. Des Moines), while employers in smaller areas usually produce the zip codes more readily. Since the employers and economic developers are interested in labor availability, they are not concerned if available labor meets the BLS definition of unemployed. They want to know their supply of interested workers, including persons who are employed, unemployed, and not currently in the labor force. The utilities companies often help give money for the surveys. This handout is just the pilot. The more counties a firm covers, the harder the guesses get. Multiples are a big problem, as well as DMV information (because their drivers licenses are renewed only every four years, and every two years when you get older). They are trying to automate this analysis as much as possible.
The following states have access to DMV: South Dakota, Iowa, Nebraska, and Wyoming
Montana and New Mexico are working on access, and John M. doesn’t know if Utah has access.
Tom G. asked if Wyoming published enough information in Trends in the last couple issues to help other states who want to do imputation, and said we probably need to meet a year from now somewhere else to walk through it. If the procedure is well enough documented in FoxPro, that, too, could be documented for the states that want to learn the program. Or, states could come to Wyoming and see how it works by running their data through. We’re hoping for some critique.
Phil B. said they’re not yet ready to be that advanced. Phil G. is not yet sure if the article had enough information to help them get started.
Bob asked if Wyoming has reviewed the Census procedures for imputation. Tom G. and Tony said yes, but their imputations are based on such large numbers that when you look at the small number it is very watered down.
Bob said that ALMIS is putting on a survey design course. Phil G. added that there are a few openings for the Portland course in a few weeks, some people have bailed out. Iowa sent some people to the LMI and SPSS design courses, and felt the LMI course was better.
IXX. State Workforce Development Perspective by Alfrieda Gonzales
The Workforce Development Council started out as the "givers of dollars under JTPA". The Council has since evolved - it isn’t a Council that just assumes role of a council under WIA, it’s broader than that. An Executive Order from Governor Geringer charged the Council with taking a broader look than JTPA or WIA. They were to look across all of the state, at all of the players, and build a strategic plan around that, in order to build Wyoming’s system. The Council analyzes what is currently taking place, their effectiveness, if they have been good public stewards, if services are currently being delivered, and how they might do things differently. Given the Governor’s and WIA’s charges, Alfrieda thinks they have all of the players there from key contributors to community colleges to economic development. All players may not be at the same place, though. They continue to struggle with higher ed in state. They also struggle with the Department of Education, which feels their role is to educate, not to supply workers to meet the needs of the state. People like Skip Gillum are leading the way, though. The Council is developing a strategic plan to fit the broader needs of workforce development. They want a strong economy, educated and trained workers, self-sufficient workers, and to close the education gap in order to meet the needs out there.
The Council is mostly comprised of business people. They do not have the pressure from a business aspect that some other states have because Wyoming doesn’t have a lot of larger/big businesses. This is a challenge because without big businesses bringing the full force of what they need to the table, government will not get excited about issues. Unless businesses demand, we may not have changes.
One of the things the Council is working towards is bringing good information to the council - including information from Research & Planning ("R&P"). They are beginning to ask questions about how the system works, demographics, where job openings will be, where gaps will be, the impact of retirement, turnover impact, and whether the workforce is productive. Because of the wealth of information, sometimes they get lost in it. So, Alfrieda’s office is trying to be very targeted about the type of information they put before them, and she includes R&P on conference calls to educate the Council members. Not all of the members have a need to have all of the information, but they need to better understand the information. They’ve made decisions intuitively in past, but now need to become educated. They need to identify important questions, find out about the data, and begin to make policies around that. Workforce Development and the Council have developed a good relationship with R&P and work with them on an on-going basis. The Workforce Development office relies on R&P to help with institutional knowledge and to help them when taking the information back to the Council. Together they participate in many planning sessions, trying to identify the kind of information the Council will need, and incorporating the information into funding requests that will go out. Up until recently Alfrieda is not sure they had a strategic plan to focus in on specific areas, and thinks it is working better now. She encourages other state LMI offices to get involved in planning with their Workforce Development offices so they will have funding resources available. She encourages the LMI offices to develop a relationship with their councils, understand where the councils are at, and work closely with them. Also, have good information for all consumers of One-Stop so the consumers can make good decisions around training, jobs, and skills to compete for jobs. Alfrieda sees the councils going beyond WIA, becoming more sophisticated, and being provided with information that they need to make the policy decisions. She encourages the LMI offices to continue to look at sharing information and doing studies on a comparable basis (compare with other states).
Tom G. explained that one thing the Council used our analysis for was to take our analysis of minimum wage to the legislature to raise our State minimum wage to the Federal minimum wage. It can work, and to some degree it depends upon our ability to develop comparable products.
Pat said Iowa gets a lot of customers that say "it would be nice to know this"; how to you target that?
Alfrieda said they were trying to answer all types of questions and didn’t know why (which is part of the education process). Something that has helped them is their strategic plan which emphasized four major areas to target. Their goal is to give R&P time when they need specific data. They were trying to bite off too much to begin with, and have now decided to begin focusing on their goal areas. Their strategic plan is different than the compliance document that they submitted for compliance with WIA; the strategic plan is more of a working document.
Pat said that it is good that they’re answering question of what information they need right away.
Phil G. added that when South Dakota does the LMI One-Stop Grant, they don’t break it down - he didn’t see the individual projects from an LMI standpoint. What Wyoming is doing seems real good. Tom suggested that if we could sit down with LMI and workforce development offices and look across the states at common features they would like to have addressed, that would help across the states. Alfrieda said that people shy away from using LMI because it’s so intimidating to them, but if they could determine priorities, issues could be addressed, it would help in the educational process, also among communication amongst states. Tom G. added that one major problem in LMI is there is no accumulation, it’s done in isolation. The fact that the three states prepared a benefits survey is great, but it would be nice to have the same reference period, survey instruments, and so on.
XX. On the Road: Where do we go from here, a review of the NASWA LMI subcommittee proposal for a wage records/administrative data program, standardized MOUs, regional goals between now and the next meeting. Content needs for an LMI Institute.
There was a roundtable discussion about credibility. Tom G. explained that we’re in sales. We have to make the links. We present ideas in Trends, then come back to a meeting and explain "Here’s how we’d approach it." The intent is to communicate ideas, not communicate a concrete finding. You don’t need to have 100% of the information you need to work with coming out of the drawer, there are ways to work with incomplete data. The Trends article on WR is intended to communicate with the public where we are at. People are requesting more and more background on WR activity. The NASWA publication has been on our Internet site ( since Tuesday (July 31, 2001). The proposal is in six chapters. The introduction lays out what the Workforce Analysis Subcommittee wants to do - have an administrative WR & 202 a standard in every state. They are proposing 1) that the WIC charter a policy council in which they have an oversight function, and 2) $20 million funding request. The funding request amounts to about three positions for smaller states and five for the larger states. Some states don’t have sufficient hardware and aren’t archiving because of limitations, while some states have staffing problems. Tony & Sylvia are doing things with WR, but that’s only part of their jobs. We’re getting a large number of domestic data sets (DMV, UI, etc.), and now Interstate data. We are compiling a large number of boxes of things that we have yet to work with. This kind of work is not a BLS manual where you’re told what to do. For national purposes we need to have standard software. We also need to be able to defend what we would do with the three to five positions in our states. There would be no transmittal of confidential data to DC, and data would turn up in aggregate form. We [the states] would have to develop these regional agreements further.
The earliest that we could get something like this done would be in the technical amendments of WIA, or within the budget process, it wouldn’t be until 2004. In the interim, meetings like this, planning meetings, should take place.
Pat said that the first thing he’d have to do would be to take ownership of these files. Tom said this is something that would be have to addressed between ETA and each UI. We also need to look at data quality. On the second to last page there is a paragraph regarding quality.
Phil B. announced that upcoming, there will be a panel with NASWA in Oklahoma City on solving WR problems; WR will be well discussed there. Executive-type people will be at that meeting.
Tom G. explained that Chapter III says that BLS strategy can be part of the solution, but CES can’t be conducted at a low enough level to answer the questions. The Decennial Census approach suffers the same problem - it’s a snapshot in time.
Chapter IV (Historic Context) asks since a WR program could have done from about 1990 forward, why didn’t it happen? In 1988 (when UI was instructed to go from Wage Requests to Wage Records), academia got real excited about working with the Wage Records. In 1992, there were amendments to JTPA with a proposal to set up a national database using WR. BLS worked on this and were supposed to do a national report to Congress within twelve months. BLS avoided delivering the report. Between 1995 and 1996, legislators started talking about a welfare reform bill which required state "new hires" databases. The only funding that takes place for WR follow-up is at the program- specific level.
Chapter V (Methodological Note) describes the differences between performance measures and evaluation measures and what the distinction is, while Chapter VI addresses confidentiality issues.
The subcommittee had more than the NASWA administrators in mind as an audience when they prepared this report.
Phil B. thinks BLS knows this [WR] needs to be done; they can benefit from it too. But it may be something they don’t want to take the lead on. They see their role as a separate entity. They wouldn’t want to be a lead, but would want to be involved. If better records comes out of there, they would benefit.
Tom G. said that employment in Wyoming is going up and unemployment is remaining constant. Are people taking more jobs or are people coming from out-of-state for the jobs? If employment in oil and gas is going up, are they claimants or are they people coming into the state? It goes back to what Tony and Sylvia presented this morning; what percentage were never employed in Wyoming before, etc. If you look at what these companies do to economic development, you have to have a product to sell. This is what we can do to explain the labor force changes each quarter. Phil B. said the Policy Council is interested in hearing about this. Tom said that one of the first recommendations is that the Policy Council issue a recommendation on a policy relating to confidentiality.
Bob L. said people need to be educated on what the uses of administrative records are. Part of your effort has to be a marketing to the other states as to what administrative records can do, and give examples of what they can do to get to WR, give them procedures. What are the five things you can get out of WR that you can’t get now, that you aren’t getting from other places?
Tom said that the LMI Institute intended to send someone to this meeting, but pulled out when they found out how expensive a plane ticket would be. Ted wanted it laid out what the Institute could offer - content needs for an LMI Institute program. Phil G. said program evaluation [follow-up] should be at the forefront of the needs, but Bob said there is a big contingency that don’t want their programs evaluated. Phil G. explained that under WIA the educational programs don’t have a choice. Bob said people need to understand the power this database could have. Show how you generate labor rates, how to interpret, what it means to the local people. Take from academic to practical.
In this meeting, we covered at the surface issues of turnover, instability index, how to store records to accommodate WR, but Tom G. said Wyoming is more concerned with conceptual issues. Academics haven’t had access to these records, so their definitions of turnover, etc., have been limited. Nobody that he knows is developing the issue of turnover - it has to be linked to issue of retention because it costs employers.
Phil G. added that it is critical to get some standardization - South Dakota does things differently than Wyoming does. We talk about some things, they talk about others, but everybody’s got different terminology. It’s important to be able to compare your data to the other states. The content has to include solutions.
Phil G. said we are creating this to solve a problem. Tom asked, "Where is the next labor shortage going to be?" We’re going to try to solve that problem. There are a lot of issues you can use WR to address. Minnesota, South Dakota, Wyoming, and Alaska all compute turnover, but we need to get together and explain it.
We need consensus from the WIC plan. What are the problems we’re facing under Workforce Development so that these problems become part of the solution? Everyone’s got someone in economic development who wants to know about turnover. Turnover is a problem not only for the employer. We need to develop concepts, not just pass around data.
Pat thinks that where turnover is concerned, some employers just want to train for openings. They ask for turnover data, but they really need the data for openings. Tom G. said some may be asking for turnover because it’s popular to ask for. A course might need to be designed around what problems the WIBs are facing.
Phil B. suggested doing a panel at an LMI forum or NASWA conference to get the internal group sold on it first, then go from there. Pat said we need to explain why you use the 202 data for one thing, and the WR data for other purposes.
Tom G. said Trends always tries to explain what the unit of analysis is. The issue of nomenclature is important; there is a need to clarify differences. People are importing language from their own discipline. For example, Tony brings the discipline of research psychology.
Tom G. asked if there were any other comments of what we should suggest to LMI Institute. Craig said he has been to a LMI panel before. He suggested getting five states that are interested and assign a lead state, then have each of the five states approach a Trends article on an assigned subject. This might serve as a basis for topics for the panel at next fall’s panel. Phil B. said that it is a two part thing. We want to "Wow" people with a new thing, then tell them "Here is an answer to this problem." They’re the ones who are going to come up with the ideas of how you can apply this in different settings. They may not know how to apply it, but.... Craig said a presentation that would "Wow" people could be tempered if you presented it beforehand. "Devil’s advocate" type thing; may get some people for it before the panel. Phil B. said people may shut themselves off beforehand if it is presented in articles - they might look at it and say, "This doesn’t apply to me."
Tom G. said the initial effort to get turnover out there was a big process. Tony’s definition of "layoff" of "x" people wasn’t really a layoff...inspiring so many other questions.... Phil G. added that Alfrieda has said turnover data by itself doesn’t solve problems; put it in terms of what Workforce Development will understand. Bob suggested if you put articles in routinely and people would pick up that you are using this data in this manner and they will want to use it too.
The idea is selling people on becoming involved on using administrative data, Tom G. said. We need some strategies to address Workforce Development questions. We’ve done a good job of selling to other states, but have to address their questions. We need to solve problems, but then also publish. What Sylvia is doing takes a long time to get to the point where we can tell the story and publish the data.
Phil B. said the LMI Institute is developing the curriculum for advanced analyst training, so it would make sense to have some short segment on WR training, say what some states are doing.
Pat was told that they were even thinking of replacing the training with specific subject area training. Craig had heard that too.
Tom G. asked what we want to accomplish between now and the next meeting.
We need to document and report out to other states; this is the problem and this is our solution. The problem is that the states participating are all small states. We need to try to get larger states involved. Tom G. also suggested identifying a common research project - demographic, industry by age, and earnings. Would like something on similar basis of what Sylvia’s working on. Phil G suggested doing something without DMV records since some people don’t have DMV info. He’s interested in turnover. Phil G. said turnover is the first stepping stone to anything. South Dakota and Wyoming are working on turnover, and need one big state. Need Minnesota or someone else who you know is working on it. Tom G. asked if there is anything we can all do together? Perhaps just descriptive statistics from the states?
The majority of states liked the suggestion of working on turnover. The target would be to compute the turnover by industry for the four quarters of 2000 using the common standard which was discussed in the first day of this meeting. South Dakota, Wyoming, Utah, Montana, and possibly New Mexico, agreed that they would like to work on this. Nebraska will not at this point.
Tom G. asked for other states to work on common demographics. Wyoming, South Dakota, Nebraska, and possibly Iowa would be interested in this. Tom G. will also send Jay Mousa with Minnesota a request. The format will be like Wyoming did in the Outlook 2000 publication in the appendix. This data will also be for 2000.
There was a suggestion to hold another WR meeting six months from now, or perhaps in March, but the consensus was to see where the states are on the above work.
|Framework for the Analysis of Labor Dynamics with Administrative Records|
|Unit of Analysis: Individual||Workforce Development System||Market||Unit of Analysis: Firm|
|Variable Type||(A)Formal Systems||(1)Industry Use||Variable Type|
|Level of Analysis||Hi Turnover Hi Retention||Level of Analysis|
|Time Space||UI||Hi Utilization Lo Utilization||Space Time|
|Education / Training|
|Individual||Hi Use Lo Use||Firm|
|Cohort Analysis||Enter from Exit to||Cohort Analysis|
|Goal/Strategy||Hi Market Attachment||Goal/Strategy|
|a living||Human Resources|
|Social Group||(minimize||Neighborhood structure||(3)Geo-community Use||(minimize cost)||Industry|
|Client Group||Community||In-Migration Out-Migration|
|UI Claimant*||Multiple job||Association/Fraternal Organ||Market Survival||Retention (pay, benft,||Client Group|
|TANF||holding||Class||Birth Death||training) Policies||Small Business|
|Student*||Job Changing||Internlocking Directorates||Recruitment|
|(Carl/Perkins)||Obtaining||Vertically Integrated Corporates||Strategies|
|Training||Earnings by Industry,|
|Population||Age, Gender||Total Market|
|*Funding for analysis||dynamic mediated by WFD (e.g Sec 122d) and market signals||Start Here|
|TNG (7/01) WYDOE|