Travis County Commissioners Court
May 8, 2007
Item 22
22, receive briefing from ray marshal center. That's not mr. Marshal himself, but chris and I worked on at least one committee together and probably more than that, but from ray marshal center at the university of Texas at Austin on finding from evaluation of county funded workforce development services with focus on the first year of the rapid employment model pilot.
>> I want to pass it over to lawrence lieman, who is our social services program administrator and has the responsibility of coordinating our workforce effort and health and human services.
>> I want to thank you for your help in this project. It helps get services to people who need it in a more effective manner. It's been a joint project all along with work source. And tamara atkinson from work source is in the audience and if any questions come up, I will make her answer them if I can't. And also to my right dr. Chris king. To my left tara smith. They've spent a lot of time looking at what to do, looking at cameras and outcomes. I am very pleased. I will at this point refer to dr. King to go through what he has to say, and then of course any further things we'll have to answer and if you need to get into any further efforts.
>> judge, yes, you chaired the social equity commission and I served as a member be and it was a great experience, thank you. So what I'll do is basically give you the kind of knock the top off the findings that we had from the evaluation rather than take you through each and every slide reading it to you. This is a project that you launched with wosh force back -- with workforce back in early 2006. It works out to $1,700 per participant over the first year. Thoots right on par. It was a partnership. The notion here was to see whether you see if you funded short-term workforce services could you cut will time people were unemployed. Could you increase the time they were unemployed, boost their earnings, get them into good jobs, something that has potential for career advancement and retention? Sue so that's the basic notion. The other key part about that is that was to focus explicitly on disadvantaged clients and not just to take anybody ta came in the door. We're trying to help do the right things with people that are particularly disadvantaged. In terms of specific outcomes, and I'm moving through the slides pretty quickly here, so expected outcomes, again, connect people with significant employment barriers to jobs paying at least nine dollars an hour thereabouts, which is a decent wage around these parts. Employment with opportunities to advance, reduced time unemployed. And for those who were coming out of the criminal justice system, also to give them the wherewithal to get back and do res ta constitution. So that's an angle that turns out we knew is more important than we thought at the beginning of this project. Let me slip to the providers.
>> there's a quick list. A.c.c. Does most of the actual provision for. Goodwill industries, the institute for child care excellence. And the professional institute of dental assisting. On the next slide you see some of the training options that were available, and just to look over those quickly, you can see it's a pretty broad list. I'll tell you in a minute. That list got shortened a little bit as the clientele changed as the project kind of unfolded and that's one of our important themes. Just to jump the punch line a little bit, it turns out most of the participants that went through here were project rio clients, so they had criminal justice experience which made some of these jobs, some of the training really inappropriate for those folks. They couldn't get bonded. Some of those occupations turned out not to work with the clientele. That's one of these evolutionary things in the project. To get ride right to the findings in trm of the process. We did both a process or implementation study and an outcome study. 112 individuals got served over the first four rounds of the project. There's a fifth round that we didn't look at, but it only had four people in and it was only late in the year. Roughly 83%, four out of five participants were out of rio. That was not intended in the beginning, it's just the way it worked out. They were also supposed to come to the food stamp, ent program frrks the welfare program, but it turns mouse of them came from rio. There's some good news findings there, but let's get to the next slide which kind of gives you the summary. Basically the good news is work source and working on this project largely exceeded its three main goals and here they are. In terms of serving clients they served 112. The target was 80. In terms of completing the preemployment training to kind of get ready to go into the longer term stuff, 79 versus 80 plants. In terms of completing the short-term occupational skills component, actually 97 compared to the goal of 64, so all in all I'd say in term of the way things unfolded, even though there were some differences, things were pretty much going at least as well as had been planned.
>> it appears that you exceeded the goals, which is good.
>> it looks good.
>> it looks real good. I'm just wondering if this is following some type after trend, especially looking at these that (indiscernible). So I guess just looking at it, is this basically that we're exceeding our goals, which is good. I don't knock it. What do you think some of the factors are that we actually exceeded the goals that you had set forces? Could you maybe hit on that a little bit?
>> I think just getting into and through the process that we designed, I think most of the credit goes to the staff at work source.
>> the staff at work source really did a great job in helping find clients that were appropriate and were reasonably committed to get through and given the support they needed to get through. And I think also one of the other parts I think is very -- I think reflects favorably on the short-term nature, be really focusing on when you have hard to serve clients, and you really minimize the time commitment so they can get in, get out and get on with working and paying their bills. So I think that's -- so I'd like to take a little credit for design, but I think it mostly comes down to the credit at work source:
>> I'll pat the court on the back also. The judge has already heard this. I think one of the unique things about Austin is a few other places, houston, Austin and a few other localities, you're actually putting local tax dollars at work in the workforce talent arena. Most rely on the federal monies that come down the pike. Those come with lots of restrictions. Here you have a project that has more flexibility to taylor things to people who really need help and I think it makes a big difference. I think our local work source staff has really put their heads around this one and good training options. And as things unfolded a little bit differently in terms of clientele they were able to adapt and still make it work. So I think it's just been a good thing all the way around.
>> in that budget cycle we set aside, earmarked for several hundred thousand dollars as far as an effort as far as work development. And I guess the question is, is any of those dollars been realized in what's happening here? Are they still earmarked for something? Can you tell me what the status of those earmarked dollars is in the budget?
>> the contracts are coming to you shortly, I hope. Part of what we asked for and will be proposing a few months ago is some additional increase in the r rem pilot. Additionally to pay for a round of training. Some money to expand our expansion in the investment in the evaluation project so we can continue taking advantage of the ray marshal center and getting --
>> utilizing, not taking advantage of.
>> [ laughter ]
>> capital idea will hopefully address essentially some gap funding as their enrollment drive is a little ahead of their funding base. It will bridge the gaps and help keep them whole through the remainder of the year. That's where we stand on that.
>> let's move to the next one. As I mentioned. We started out with eight training options. That dropped to three by the third round. And basically that's a function of having the rio clientele be the dominant group being served. So mainly administrative assistant, truck driving and construction. Definitely that clientele was a factor in who we could serve be and serve appropriately in those training options. The other thing to note about that, and we've had these sctds with work source and with lawrence and the staff is that brought its own challenges. There's some mismatches between scheduling with the probation and parole people. They want meetings during the day. Sometimes they yanked people out of training. Well, they're supposed to make their training. So there's some -- I guess a little bit of rubbing at the edges on some of those things, but the friction is being worked on. Probably there's more that can be done in that arena if in fact the clientele remembers to be predominantly a rio clientele. But everybody is aware of that and paying attention to it. Another finding from the process side, they were offered support services as they went through the training, but as you look across those dividers, the kind of assistance they got and the intensity of those services varied a bit. And finally, placement services was a big part of this design. There was a thoughtful structure, kind of a sequence of events that was going to happen. It didn't quite unfold that way, but in fact be most of the participants were getting some job be search assistance as they went through from their providers. Some of the glitches have been worked out, so I think it's a bit smoother now in the program that's underway. So if I could shift to kind of the outcomes side of that. We're looking at different data sources. On the one hand we started by looking at data from the providers, looking at some of the reports that came into the county and work source. So the first reports were on that basis that about 65% of the 112 that went through were employed either full or part time, immediately after they finished participating, making a wage that was over 10.60 or so an hour. Remember the goal was nine dollars an hour, so based on those provider reports, things are looking pretty good. We also -- we also looked at if something called ui wage records. We basically take identifiers from the participants, we take them to workforce commission and we get their employment and earnings that employers report. We like these because they are independent, they are pretty solid, they're audittable. Necessity don't have a buy as -- they don't have a buy as. They want them to look higher. So you've got a tug in both directions. Generally speaking we found in 25 years these are pretty good data sources to go to when you want valid outcomes data. You don't get hourly wages or weekly wages or annual salary, but in any eye vent, this is what we found. Looking at these initial ui wage records that we could grab, and there's a little time delay, basically if you look at that third column over, we took the average of the four quarters before they enrolled in the rem project and you can see something like a fifth of these poom were employed-- people were employed during that average of the four quarters. So it's a pretty disadvantaged population. In that last quarter, where they're getting service with rem, the numbers are jumping up and that last column is basically a second quarter after they've left service. And that's the most recent one we could get data forgiven the time lagz in the data. So basically if you look at the bottom row there. So overall in the four rounds for which we have some data, 18 and a half percent employed before, more than half, almost 53% employed two quarters out. That's a trip willing almost quaird to their -- tripling almost compared to their preprogram. That's a sampling. I should tell you the next faye of the work we're doing we have to do more rigorous records to look at the net value added in their participation against some comparison groups. Not doing a random assignment thing, but we'll have a bitter way of attributing the impact be of the rem peangs. But these are looking I think quite respectable. It's a little lower than the provider reports, but that's probably to be expected for a couple of reasons. One, there's some under coverage issues in insurance and construction is one of those areas. That's one of the fields these guys have gone into. And truck driving.
>> [one moment, please, for change in captioners]
>> .
>> as chris pointed out, the providers were basing employment on people showing an astub. In as much as a lot of our participants went into construction fields and truck driving. We know those are fields not covered very often by un unemployment insurance, so they would not have wamings reported to the ui system, and that could account for some of the lower findings in the ui record.
>> it looks like the ui records are showing a larger spread for services. I'm trying to figure out.
>> may well be.
>> the differences between the table.
>> I don't think we have a great way to account for differences in that. What I can tell you is that over time, we always revert back to the ui system because it's auditable and both sides have more confidence in the re reliability. Over time the trend auto to look about the same, we would hope. If I can shift to the earnings, then, to finish out the outcome picture in the next slide. You can see there basically what happens with the pre preprogram, end program and post program earnings. The picture is generally up upward, as we would hope. Generally speaking if you are participating in a training program k your earnings tend to dip a little bit. We see that because people are hopefully paying the attention to class and not working quite as much doing that. The other thing, we have about a 70 percent increase overall between the preand post period, up to about 35 3500 per quarter in earnings earnings. And that is quite respect respectable. The other thing we watch for over time, and we have seen this from a national evaluation and other work that we have dan at the ray marshall center, it takes a way for the training to take kick in and get the full return sometimes that might be a year and a half, two years down the road. I'm not sure whether we will be able to do it quite that long, but again, we will track to see whether we are getting the upward trend and come back to you, and we will have the value-added piece to add on to that.
>> that is a very exciting thing, I think.
>> yeah.
>> the v?l-added piece.
>> not too many people at the local level are able to do that.
>> that is one thing for us to measure our own performance against the expected outcome, and another to see what the penetration is against a control group, essentially, to see what folks would be making even without that. So valuable.
>> there are some limitation limitations to doing that and we are looking into that that. If you don't have good demographics and other data from the providers, it's hard to get a good match on a comparison group. So we are again working with the providers to kind of make sure the data systems are up to snuff and give us everything we need. It's one of the niches that the ray March shall center has pioneered over the last 20 years, making use of those data and using them for program improvements. This has been a great opportunity for us as well. We have typically done it statewide, not is much at the local level.
>> does the fact that the service population ends up being mostly, does that impact the controls.
>> it makes it harder actually. And here is the reason. What you like to do, you want to test for selection bias in the evaluation. Do we just get people more motivated. The problem is you want to test using preprogram earns. For rio people, almost by defer nigs they didn't have so much reported earnings in the preperiod. They were busy. So we are going to try that out. It's going to be a stretch for us but we will report back to you it is also harder for very young participants because they haven't settled into a nice career path yet. For those two populations it's more difficult. But again, we like challenge challenges.
>> one advantage also, an opportunity we have (inaudible) the workforce and the stin also contracted with the center to do similar evaluation looking at the ui wage data for those traditional workforce programs that the city and county have contracted with over a number of years. So we have something, our pilot, the innovative approach compared to just the baseline of what has already been here, in addition to the comparison group that the center is putting together. So we have I think, mope mope--hopefully a really notice picture of what is happening.
>> on that score, we are actually going back to 2002- 2002-2003 participants so that for those who finished early enough, we ought to be able to look at 2-3 years out and see, again, some of that return kicking in. We are doing the same kind of quasi experimental design design. It takes longer to get a post track but we have been talking to steve jack jack o obs and his staff. In items of conclusion observations, the news is basically good at this point point. The rem model largely met its participation and training completion goals, the initial employment goals the wages and entry targets, earnings track looking as we would hope it to look. Even though we had had some rough spots in here with the job search piece and the referral mechanisms, we talked with staff about those and they are already working on those things. Similarly for the training offers, they adapted very quickly and we have been giving some thoughts to that with them for the rounds that are now going on. The one that probably needs more work, the next couple really, improved coordination with the parole system to work out some of those glitches. If rio people are going to be who we are dealing w whee some thoughts about that. Also, improved provider reporting of their data for participants, and to some extent their outcome. In terms of next steps, we will add more data to this, round out that longer term picture, give you the vad- vad-added piecepiece, and work with you to improve data collection and reporting. We would be happy to answer any additional questions you might have.
>> explain to me again how we ended up with such a large concentration of ex exoffenders.
>> you know, I'm not sure I really know the whole story on that one, but I think there's been good outreach towards the exoffender population.
>> okay.
>> the other thing, I think some of the choices people and the food stamp clients have more mainstream options through work source. They have regular funding streams that flow through there, would be at least two factors I would think of.
>> would you also say maybe the flexibility of the funding as well, compared to some of the more rigid streams that workforce has had to operate in ?
>> I'll accept those words. I think that is right.
>> cameron is nodding in the audience. The flexibility that we offer is the primary thing to address the biggest thing that the federal funded system created.
>> I don't have any problem with that. If you accept that as reality and say okay, we have more than our share of exoffenders anyway, and here are three job categories that they work in, what job categories would you add? It really kind of turns on employer willingness to hire offenders.
>> sure. I would think there are probably some other jobs out there that we haven't maybe fully tapped. There are some things probably in warehousing. If you think of the ones that are booming, construction is one, and truck driving am we are doing a lot of shipping and related work as the economy continues to baffle those of us that watch these things, and grow and grow. I don't know. Warehousing. We haven't explored other occupations, at least I haven't. Work source is probably looking more carefully at that. We are working with work source at some of the high end jobs, connecting with the st?n Austin and chim ber initiative. You will have to jump up to technician level, I think. This is short-term training of some of that will involve lengthier training. This is really to focus on a particular need for not many short-term training options that give you good wages but certainly something we could explore.
>> our interest from the beginning was to make folks on this a first step into the job market with the idea a lot of resources exist in the community if you have at least a decent foundation from which to start. For example, if you have little or no earnings, no familiar or support, criminal record, if you are under the gun to get into something right now, capital might be a viable option, but as a first step, if someone is successful, they can get a solid footing in construction, in the administrative trains, then maybe a capital idea is an option or going back to acc on their own or any other option in the community. This is a step one to get sufficiently stable so that it's, you know, they can pay their basic fees, get housed some way on their own without depending fully on public assistance.
>> laurence's mentioned acc and I think that is important. Acc is really the main provider of training in the whole reasoning. For many of the folks coming out of rio, I would expect they have not had much contact with the post secondary training institutions. The more you connect them up the better off they are going to be because they find out that it's not a foreign place to go. You can go back for life long learning, add a credential or two, probably not a bad start.
>> part of what work source does on a constant basis is monitor the job market, see what is happening and emerging and changing, going away. The advantage of having, one of the many advantages of having them as a partner, as stuff comes up they are also looking for people to train and meet those needs and in doing that we can see if we find training, jobs and training amenable to the population, we have the latitude to bring those into the network and continue to expand and diverse fy the opportunities that we have where feasible.
>> if you wanted to identify the, I guess, communities that are representing the model for best practices by hiring exoffenders, the real hard to employ persons, any communities stand out?
>> that's shameless, judge. Of course Austin stands out. Seriously. I do think Austin is a model for trying to do this. The ones that come to mind as well would be kind of the bay area, new york city, and new york has some of the old oldest projects run by the wild cat corporation and others, philadelphia. Honestly, I would put up Austin's workforce providers their city and county folks and others, against the best of them.
>> can you do this kind of penetration analysis on every one of the divisions in the county?
>> I am not sure it would be good to be used.
>> take advantage of it that way.
>> we are doing some other work around the state that we are just kicking off to look at return on investment from other workforce investments. And sphthat takes takes--and if that takes off, Austin will be included on that, kind of the broad based returns on investing in workforce services. It will be interesting to see how that unfolds.
>> to compliment, you just paid us a high compliment and I want to pay one right back. If anybody is interested in looking at what a top-notch product on the performance side the penetration analysis looks like, this is just great.
>> thank you.
>> really, really helpful and useful.
>> I should tell you, we are also tracking high school seniors in the central Texas region, tracking about 70 percent of the graduates in the ten-county region, 95 percent in Travis County. This is, again, unfolding, longer term, but tracking them into post secondary education training and into the workforce. And this is a multiyear project again, those reports will be on our website. We have some briefing materials. We surveyed the seniors as they exit each spring. You hear tara coughing. She is exhausted from going to the high schools. We connect up with the administrative data and get a pretty good picture of what is happening to the high school kids. So again, we are trying to kind of look at all of the human resources and how the inputs are working for the market and for post secondary.
>> that's great.
>> yep.
>> judge and Commissioners, I will just close by saying our thanks to you because without the opportunity for staff to be creative and get with our partner to pull this project together and then begin with an evaluation such as the one you have just received, we would not be anible to provide you with these out outcomes. Also you are your investment has brought in additional investment from both workforce and the city of Austin.
>> thank you.
>> I told you we were on the cutting edge. You didn't believe me. Thank you very much. Good to see you again. Thank you both of you. And staff.
>> thank you very much.
The Closed Caption log for this Commissioners Court agenda item is provided by Travis County Internet Services. Since this file is derived from the Closed Captions created during live cablecasts, there are occasional spelling and grammatical errors. This Closed Caption log is not an official record the Commissioners Court Meeting and cannot be relied on for official purposes. For official records please contact the County Clerk at (512) 854-4722.
Last Modified:
Wednesday, May 9, 2007, 8:00 AM