Employment Connection: Demand-driven model increases job placement
April 3, 2013
Download summary (2 pgs) Download report Press release
By implementing a demand-driven model in mid-2010, Employment Connection, the workforce-training agency for Cleveland and Cuyahoga County, has greatly improved the services it provides to local firms and workers.
Employment Connection is the combined workforce delivery system for the city of Cleveland and Cuyahoga County, serving Ohio’s Local Workforce Area 3. In the middle of 2010, the agency implemented a demand-facing service delivery model that focuses on building relationships with regional employers and linking them to pre-screened, qualified job candidates.
As a result of this change, more of the agency’s clients have been finding jobs, an increase in both the overall number of placements and as a percentage of total clients served.
Key findings include:
- The share of clients exiting the agency’s programs who have found jobs has nearly doubled, growing to 84.4 percent in 2011from 43.6 percent in 2009;
- More than five times as many exiters are employed, increasing to 2,470 in 2011 from 471 in 2009;
- Although Employment Connection now emphasizes job matching and placement, the percent of exiters who have received training is on par with previous years, with slightly more exiting without training in 2011, and fewer exiting without training in the first eight months of 2012;
- The demand-facing model emphasizes employer-connected, shorter-term, on-the-job training, The share of exiters accessing on-the-job training grew to 12.5 percent in 2011, from just 0.8 percent in 2009;
- Earnings have rebounded for adults and dislocated workers exiting agency programs, from a low point in 2009, and are closer to earnings reported in 2008.
These improvements have happened even as the agency’s budget and staff have shrunk in a period of relatively low regional job growth. The increase in placements does not appear to be the result of a shift in the educational attainment level of exiters. While the share of exiters with no college experience is smaller in 2011 than in 2009, the overall number of exiters with no college experience more than doubled, and the placement rate for this population has nearly doubled.
Employment Connection’s outcomes have quickly improved after implementing the demand-facing model of service delivery. While the program’s improvement has been remarkable, the most important lesson from the EC transition is not necessarily the model the agency is now using; rather, it’s the idea that the service-delivery model must be closely tailored to the needs of the agency’s clients, local employers, and the regional economy.
Diversity is one of the strengths of Ohio’s economy. The state is home to several urban areas, each with distinct labor traditions, and to rural regions that range from agriculture to Appalachian Ohio’s extraction-based economy. It should go without saying that what works in Cleveland will not necessarily work in other areas of the state. The process of building relationships with employers will be shaded by the local employer composition, and by regional and local cultural norms. Effective employer outreach in Cleveland will be very different from effective outreach in Portsmouth, but it should be happening in every Workforce Investment Act area in the state.
Evaluation and change are not easy to tackle, and if areas do not have relationships with workforce stakeholders, including employers, labor, and advocates, they may not have enough information to make good decisions about service delivery. The state could play a larger role in helping local areas bridge this relationship gap.
Sector partnerships function at the regional level and bring together workforce stakeholders, including employers, labor, training providers, the government, researchers, and community organizations to assess and address the workforce needs of specific industries. Such efforts have been linked to increases in employer productivity and to decreases in client poverty.
These partnerships can create the regional infrastructure needed to support the deep relationship building Workforce Investment Boards must undertake if they want to create closely tailored service-delivery systems. The state should use some of its workforce training funds, enlarged now through casino revenues, to support regional sector partnerships. Unlike training vouchers, which benefit individual employers and are often geared toward meeting immediate, short-term training goals, sector partnerships would support regional sector-wide productivity. Such funding would provide competitive grants to regional consortia for the development and implementation of employer and worker responsive curriculum and training. Grant criteria should prioritize partnerships that bring together a diversity of stakeholders, such as employers, training providers, workforce and community advocates, and labor and worker representatives, as well as partnerships that focus on building career pathways for moving low-skilled workers into higher-skilled jobs in in-demand occupations.
Employment Connection’s transition also highlights the importance of funding. Funding cuts have meant fewer overall staff, and even though the agency is functioning with greater efficiency, increasing placements and the number exiting, Employment Connection is serving fewer total customers. With adequate funding, the agency could expand its service capacity, increasing the number served and maintaining both its traditional and on-the-job training programs. The choice would not be driven by funding but by which program best meets client and employer needs. Efficiency is essential stewardship of the public dollar, but it is not a replacement for adequate funding. Federal and state policymakers should take a balanced approach to public budgeting, restore revenue, and avoid further restrictions to this important set of services.
Employment Connection: Demand-driven model increases job placement
Employment Connection is the brand name of the combined workforce delivery system for the city of Cleveland and Cuyahoga County, serving Local Workforce Area 3. The joint project is staffed in part by the city and county, but also by the United Labor Agency.
Until mid-2010 Employment Connection, operated in a way typical of such organizations, largely following a social service model focused on meeting individual job seekers’ needs and providing orientation, skills assessment, workshops on resume writing and job interview skills, and funding for more intense job training through individual training accounts. Many of these basic skill-building programs remain, but the overall focus has radically changed. Since mid-2010, the primary focus has been on job placement and meeting the human capital needs of local employers.
In this report, we examine the transition to a demand-facing, employer-driven model that focuses on the transition impact to the adult and dislocated worker populations served by the Workforce Investment Act. By looking beyond the Workforce Investment Act common measures, we found that the number exiting the program to employment increased five-fold. Even as more people exit the program, a higher percentage of exiters are being placed in jobs. These are surprisingly strong results in a still-troubled regional economy.
Cause for change
As Cleveland struggled to shake off the 2007-09 recession, the leadership at Employment Connection was becoming frustrated by the agency’s performance. Too many people were entering the program, signing up for training, and leaving without a job. Too often, training was not connected to employment.
The agency was concerned about the way it was perceived, that some saw it as a place to get access to financial help to attend community college or training rather than as a place to find and prepare for a career. Students were showing up with completed intake forms requesting individual training account vouchers to cover community college costs when there was no connection to an employer or completion plan. Training and skill building had always been part of EC’s programing, and the Workforce Investment Act (WIA) prioritizes training and increasing worker productivity, but EC leadership feared that the training-account component was losing its connection to work.
Beyond this uncertainty around the mission, the agency’s leaders also felt that the model was not the right fit for the labor market. Prioritizing long-term training generated different outcomes when unemployment was low, but demand for workers high, as when WIA was authorized in the late 1990s. At that time, employers were experiencing a genuine skill shortage and investment in training was often quickly rewarded with employment. Today the job market is already flooded with a labor surplus, particularly with dislocated workers who have extensive experience and work histories.
The labor market Clevelanders have faced since 2007 is loaded with challenges. Ohio has had more than three years of recovery since the 2007-09 recession, and in that time added 135,600 jobs. While this is progress, it is cold comfort, as more than 400,000 Ohioans are officially unemployed. Even with so many workers dislocated, articles often cite employers saying they are unable to find qualified applicants for open positions.
WIA reporting from Program Year (PY) 2009 reflects this struggle. In PY 2009, Employment Connection failed to meet three common measures: percentage of adults entering employment, percentage of dislocated workers entering employment, and the percentage of youth entering employment or education. Table 1 shows outcomes for adults and dislocated workers.
According to the Annual WIA report for Area 3, Employment Connection served 6,142 adult and dislocated workers in program year 2009. This was a 57.2 percent increase in the number served from the start of the 2007-09 recession in program year 2007. The agency needed to make gains in job placement to meet this demand and they needed to do it with significantly less funding.
From program years 2009 to 2010, EC’s WIA formula allocation for adult and dislocated workers fell from $8.8 million to $6.9 million. Table 2 shows the agency’s share or WIA funds by year, for adults and dislocated workers.
Except for the much-needed funding boost delivered by the American Recovery and Reinvestment Act, the state’s overall WIA allocation to adult and dislocated workers was shrinking. Area 3’s allotment is shrinking too, getting a smaller share of the shrinking state total each year.
Employment Connection needed to improve performance. Longer-term training was costly in both time and resources and was not necessarily being rewarded with a job. There were just too few jobs available for those seeking work. The model that focused on needs of clients rather than job matching, job placement, and employer relationships was not working. The agency needed to craft a delivery system that addressed the mismatches in the labor market in order to move clients into jobs.
In response to these twin pressures, agency leadership chose to restructure and embrace what they called a demand-facing service delivery model. This approach prioritizes relationships with local employers and makes meeting employer needs top priority. EC was to operate more like an employment agency. The focus of the agency would be on recruiting and building employer relationships and making successful job placements. This was a substantial shift away from the social-service approach that focused on providing assessment, training assistance, and case management services to job seekers, which had generated only 1,050 employer contacts in the preceding 10 years. To make the transition, Employment Connection needed to change both policy structure and agency culture.
Part of the solution included a reassessment of the agency’s own skill composition. The funding situation demanded staff cuts regardless of the overall priority shift. Before the transition, Employment Connection had 80 full-time equivalent city and county workers. Now, there are only 25. With fewer staff overall, the agency used federal stimulus funding to bring on staffers with deep experience in sales and recruitment, rather than staffers with backgrounds in social services. EC formed a team of business service representatives and made the team the hub of the agency.
Rather than waiting for employers to contact the agency for assistance or to list a job posting, the Business Service Team consultants use their sales skills to actively engage and promote EC human resource services and solicit job leads. The mission of the team is to help employers compete by helping them meet their human capital needs. The team promises to generate leads on well-qualified candidates for job openings within 72 hours after an employer makes the request. This creates job placement opportunities for EC clients, and helps solve the skill mismatch problems reported by employers.
Originally, the agency used five business team consultants to cover four geographic areas, with one team focused on “green” jobs. With the expiration of a green-jobs grant, and with an increasingly unbalanced workload, the teams were reconfigured. The agency now runs three larger teams, each covering a separate geographic area.
The agency considered structuring teams around industry sectors. Much of the agency’s work in employer relationship building centered on the health-care sector, in part because of successful relationships with the Cleveland Clinic and in part because health care plays a large role in the local labor market. The agency felt that dividing work based on geography was a preferable structure so that resources would not be tied to one industry but could address the needs of a variety of firms.
The reconfigured teams allow each business service consultant support from three recruiters. The consultant serves as the bridge to local employers, marketing agency services and bringing in job leads. Recruiters identify and pre-screen candidates and make the employment match.
The staff at Employment Connection helps all One-Stop clients develop a resume that is loaded into their database. The goal is to make the database and candidate search the most robust in the county so employers consistently return for human resources services, but recruiters also scour Temporary Assistance for Needy Families and unemployment rolls, as well as online job databases, such as Ohiomeansjobs.com, Linked-In and O*Net for possible candidates. They try to make the best possible match and case management is a large part of the agency’s turnaround. The system allows staff to track client and employer placement histories to document and build the history of the relationship. Once a list of candidates is identified, the agency uses robocalls to notify potential matches of the job posting and conducts pre-screening interviews by telephone. As consultants generate job leads, recruiters use this system to make skill matches and refer pre-screened, qualified applicants to employers. The pre-screening and assessment provides value to employers, who can rely on One-Stop and avoid sifting through hundreds of applicants in electronic or on-line databases.
The business representatives do more to help define their work as employer-focused and not just a more aggressive or proactive placement team. They work to convey why these services matter to employers, explaining that running hiring needs through the EC system can reduce turnover and hiring expenses. Business team members help employers identify training needs and link them to on-the-job training resources. The team follows up on placements to verify that the hire is a good match. The marketing of EC services and follow-up helps employers understand that the One-Stop is a place to find well-qualified employees, helping to erode the social-services or charity stigma that is unfairly attached to the WIA system and other public programs.
Job placement improves
Since implementing this system in July 2010, the agency has dramatically improved outcomes for its clients. According to WIA annual reports, EC’s adult job placement rate improved from 53.1 percent in program year 2009 to 82.6 percent in 2011. Similarly, the placement rate for dislocated workers improved from 62.3 percent to 87.2 percent. Area 3 went from underperforming in 2009 to exceeding all nine WIA outcome measures in 2011.
The first quarter of 2012 captures the first set of outcomes generated entirely after implementation of the new model. Table 3 provides a summary, with more detail in Appendix A.
Limitations of WIA outcome reporting
While WIA annual reporting provides some measures to gauge performance, it does not provide detail on clients, placement or training. Workforce Investment Act outcome data is also notoriously easy to manipulate. If the provider places more clients, a real positive outcome, it increases the top half of the fraction, so it correctly looks like a larger share are being placed. But outcomes can also appear to be improving by shrinking the lower half of the fraction, simply not counting some participants who didn’t get placed.
As Table ** shows, there is also tremendous lag in the data. The number entering employment is reported in the first quarter after exit. Retention and earnings are measured in the second and third quarters after exit, respectively. So retention and earnings reported in the first quarter of 2012, for example, cover those who left the program from April to June of 2011. This lag makes it difficult to identify the factors impacting performance change, or when the change occurred.
For these reasons this report relies not only on the WIA outcomes data but also the EC exiter data. Employment Connection provided exiter data from Jan. 2009 to August 2012 for this report. To understand how agency performance was changing across various measures and populations, we sorted clients based on the year they exited the program. EC had only had a couple of years make the transition, so there was limited data to use in making comparisons.
Job Placement according to EC Exiter Data
Even with the limited data available, it is clear EC has made tremendous improvements in job placement. Figure 1 shows the number of placements per year.
The number of individuals exiting has remained high and the percent finding jobs has dramatically improved. Even with incomplete data for calendar year 2012, the number exiting the program and the number exiting to employment were higher than in 2009. From 2009, the last full year under the old model, to 2011, the first full year of the new model, the number exiting to employment increased by more than 400 percent. The rate of participants exiting to employment more than doubled.
Employment Connection faced a very challenging year in 2009, with outcomes that were not as good as previous or subsequent years. The performance outcomes reported were generated during a difficult city-county merger. There were large disruptions and changes in staff at the administrative level, and a changing funding outlook. This reflects an agency struggling with those challenges and may in part be due to the structural upheaval, rather than a reflection of agency policy.
Some of the agency’s rapid turnaround could be attributed simply to a calming of the administrative waters after the merger. However, the transition to the demand model brought its own set of administrative and staff changes. The size of the agency was cut nearly 70 percent, job descriptions and organization priorities were changing, and funding was shrinking; yet even under these structural changes the staff greatly improved performance, far exceeding 2009 performance, and also exceeding outcomes from 2007 and 2008.
Employment placements improve as Cleveland economy struggles
June 2009 marked the official end of the 2007-09 recession. Ohio has been in a long slog of recovery since, and some of Area 3’s improvement could be related to the overall modest improvement of the economy. The state has made some slow gains, and Cleveland’s gains have been even slower. At the end of the 2007 recession, while the state unemployment rate dipped to 9.0 percent, Area 3 had an official unemployment rate of 9.7 percent. In 2011, the annual average fell to 8.1 percent, and continued to fall to an annual average of 7.3 percent in 2012. The improving unemployment rate, while positive, does not present a full picture of the region’s labor market – changes in the labor force can create the appearance of improvement, while underemployed workers are not captured at all.
A better measure is provided by the Bureau of Labor Statistics’ Current Employment Survey (CES), which surveys firms to estimate total number of jobs. This survey also suggests slow improvement in the job market. From the end of the 2007 recession through February 2013, the Cleveland MSA added only added 28,600 jobs. Figure 2 shows the total number of seasonally adjusted nonfarm jobs for the Cleveland MSA since the start of the 2007 recession.
According to these measures, Employment Connection’s demand-facing model appears to be maintaining a high placement rate in a tight labor market, and the agency is improving performance during a period of slow job growth.
Reorienting the agency and remixing the skill composition of the agency’s own workforce helped change its culture. With new skills added to the mix, the change was more than structural. EC leadership believes that through these activities the agency will become a trusted partner of employers. Prioritizing business services and restacking the skill deck sparked not only an internal culture change; it also led to an external perception change that is evidenced by the improved placement rate.
Agency improves its training policy
To solidify the change and accommodate the increased cost of the more robust employer services section, the agency’s approach to training had to be reevaluated. Employer services, which used about 10 percent of the agency’s budget, would require more funding to support the work of the business teams. The local board reallocated funds so that half of available funds were dedicated to identifying and responding to employer needs. The decline in WIA funding compounded the need for rebalancing. Table 4 shows how the loss of WIA funding impacted different elements of EC’s programing and how the agency flipped its traditional funding model to free up support for employer services.
Employment Connection faced a $3 million cut to WIA funding between fiscal years 2010 and 2011. This 20.9 percent reduction in WIA funding was spread through the agency’s programing in a way that reflected the changing priorities. Occupational skills training which funds the agency’s individual training account vouchers sustained the largest cut of more than $1.1 million. This was nearly a 70 percent reduction in funding. Youth training programs and job seeker services also sustained sizable cuts of more than $640,000. Supportive services, which help clients reduce barriers to completing training by helping to cover childcare and transportation costs, was cut in half. Employer services, however was left stable though it was still only slightly more than 12 percent of the total Workforce Investment Act budget. Table 5 shows how EC used additional non-WIA revenue to backfill budget losses, and details the amounts and sources of the non-WIA funds.
Employment Connection brought in enough revenue to replace 72 percent of the total Workforce Investment cut. Core and Intensive services were backfilled. Overall, the training budget was 47.8 percent of WIA funding in fiscal year 2010, 42.5 percent in 2011, and a smaller slice of the 2011 all-funds budget, 41.1 percent. In fiscal year 2010, EC allotted 11.5 percent of its WIA budget to individual training accounts; this was nearly a quarter of all WIA training dollars for the fiscal year. After the restructuring, ITAs were budgeted only 4.8 percent of WIA funds, and although ultimately other revenues were used to boost the account to 6.2 percent of the total budget, individual training accounts took a backseat to customized and on-the-job training, accounting for only 15.1 percent of the all-funds training budget. This funding change, which prioritizes the training programs that are most closely tailored to employers, solidified the demand model in the agency budget and forced major changes in how the agency supports access to training.
Individual training account change
WIA is not just about job placement. The Act also seeks to increase the occupational skill of participants, to improve the quality of the workforce, reduce dependency, and to enhance productivity and competitiveness of the nation. These directives differentiate the WIA system from private employment placement agencies and can be the source of many of the concerns about the demand-driven model.
Individual Training Accounts (ITAs) are the primary way jobseekers secure training through the WIA system. These accounts function as a voucher program; local areas set their own policies for issuing these training accounts. Most areas limit job seekers to a capped amount and capped training time, and require some participation in skill assessments and counseling designed to help the seeker make choices about training options. Some programs may choose to only provide ITAs after jobseekers complete core and intensive services without a job placement, although the U.S Department of Labor does not require that WIA services be tiered. Locals also limit job seekers to selecting training from an eligible training provider list and some limit training to in-demand occupations or industries. Ultimately, WIA requires ITA policy to “maximize customer choice” and the final decision on training is left to the job seeker.
Prior to restructuring, Employment Connection’s individual training account policy capped training at 104 weeks and $10,000. Additional requirements were added to the program to bring the training policy in line with the overall shift and “due to limited funding in the WIA formula adult and dislocated worker allocations.” In the demand-facing model, two primary restrictions were placed on ITAs: Adults are only approved if an employer is committed to hiring the individual after completing training; and; Dislocated workers can be approved only if the training will not exceed a year, the ITA is $6,000 or less, and it results in full program completion. Individual training accounts meeting those criteria are further restricted:
- They are only issued for in-demand occupations, or where a certified employment opportunity exits;
- Training must be directly related to employment placement;
- Training cannot exceed two years;
- The client must seek Pell or other student aid, as appropriate; and,
- The training account may be rescinded if the client fails to maintain a C average or maintain good standing with the training institution.
The individual training account policy supports the demand-facing orientation of the agency by prioritizing the connection to employment and limiting ITAs to relatively short-term training. The more restrictive policy also frees up resources for the business service team program and employer-directed training. This change is not without controversy.
Employment Connection’s individual training account policy could be viewed as exporting the “work first” policy stance of welfare reform to the WIA system because employment placement takes priority over training. Moving customers to work, and not necessarily through the various training or education systems is the primary goal. Some criticize this approach, citing concerns that it limits much-needed and already-limited training opportunities for low-skilled and low-income workers.
It is difficult to assess the effectiveness of training strategies in the WIA system. The WIA common measures are focused on employment, retention, and earning. Employment Connection provided exit data for calendar years 2009, 2010, 2011, and much of 2012, and included whether the exiter received traditional training (through individual training accounts), customized or on-the-job training, or no training. Figure 3 shows the share of exiters participating in each type of training program by calendar year of exit.
The majority of clients exiting do not receive training; this was true in 2009 and 2010, before the agency switched models. The share of exiters that received no training was up in the first full year of the transition by 2.7 percent from 2009 and 7 percent from 2010. However, the 2012 data show a decrease in the share of exiters leaving without participating in training, dropping below both 2009 and 2010. As the 2012 data is finalized, the additional exiters may drive the share higher or lower. It appears that training is still taking place at rates similar to pre-transition rates.
A clearer trend is seen in customized and on-the-job training participation. The share of exiters participating in employer-driven training has grown since the transition by more than 12 percent. As expected, traditional training through individual accounts has dropped. From 2009 through the first full year of the demand-facing model, the share of exiters participating in individual training accounts shrank by more than 14 percent. The 2012 numbers look more balanced, but again the data set is incomplete.
The gains in customized and on-the-job training are more pronounced when the number of exiters is examined. Table 6 shows the number of exiters by training type by exit year.
The percentage of customers accessing training is roughly on par with prior years, but when the percentage is peeled away, we find that as the total number of exiters has increased, more individuals are passing through the agency without training. The number of individuals receiving employer-connected training dramatically increased in the first full year of the program, 2011. Data through August 2012 has already surpassed customized training levels for both 2009 and 2010. The Ohio Department of Job and Family Services has noted this progress, stating that improving uptake of on-the-job training and employer-connected training has been a goal for the state system and Area 3 is “ahead of the curve.” Restoring WIA funding could help funnel more of the participants into employer-directed training.
Employment placement rates have improved across all training types. Figure 4 shows the rate of employment placement for each training category.
After a dip in 2010, traditional training placement rates have improved by more than 32 percent since Employment Connection implemented the more restrictive individual training account policy. The more open-ended approach that supported greater client discretion was not generating job placements. In 2009, half of those using individual training accounts were not exiting into employment. The customized training placement rate also improved, even while the training category was expanding. Placement rates for clients not participating in training also got better, going from the worst rate of all categories, to the best.
Exiters see increasing earnings
Improving employment placement is a positive step forward for the agency, its clients, and the community it serves. However, it is important that the Workforce Investment Act system remain a resource for improving skills and opportunity for low-skilled, low-wage adults. There are very few training options accessible to that population. Individual training accounts, particularly structured programs that require jobseeker guidance and counseling but provide larger awards, have been shown to result in greater earnings and public savings over time than programs that restricted awards and capped time.
If the demand-facing policy were resulting in shrinking opportunities for participants, or shuttling clients into work at the expense of earnings, Employment Connection would post declines in average exiter earnings. That’s not the case. Figure 5 shows average earnings by year according to WIA outcomes reporting.
EC is making gains in exiter wages, not losing ground. The 2011 outcome reports shows higher average earnings for both adults and dislocated workers leaving the program. Earnings have returned to 2008 levels in terms of actual buying power. There is also improving parity between exiters in the two programs. Area 3 exiters are making gains on the area average wage, but fall slightly below 2008 levels. Table 7 shows exiter earnings by year, as a percent of the average Area 3 wage.
Training is still taking place in the Employment Connection model and exiters are closing the gap between program earnings and average earnings for the region. The agency’s 2011 earnings data is much closer to the area average in comparison to the two prior years, but it still lags behind the 2008 numbers. The six-month average wage for Area 3 in 2008, in the middle of the recession, was $208 less than the average in 2011.
On-the-job training policy promotes quality
Employment Connection’s move to prioritize on-the-job and employer-connected training also required a revamped on-the-job training policy to ensure quality. The agency’s on-the-job training program provides employers reimbursement up to 50 percent of wage rate of the trainee, and extraordinary costs of training and supervision. EC offers a variable reimbursement scale for smaller employers. The maximum on-the-job training amount is $5,000 per trainee.
The agency’s policy ensures that on-the-job training provides benefits to the trainee, and goes beyond subsidizing hiring wages for employers. The training must provide the skills and knowledge for full and adequate job performance, and it must be limited to a time reasonable for the particular occupation. Employers must state that they intend to hire the trainee at the end of the contract. The position must be at least 32 hours per week, cannot be temporary or seasonal, must pay at least $10 per hour and should have career advancement potential.
Basic accountability and eligibility screening guidelines can help make the on-the-job training experience a positive one for employers and for workers, and ensure that public dollars are not subsidizing low-wage, dead-end job placements. While the actual training placements were not reviewed, the earnings data and the policies of EC suggest that moving to more employer-connected training has not resulted in poorer outcomes for clients.
Who is being helped?
Employment Connection’s placement gains should also be looked at in light of the agency’s client population. Exiting higher numbers of clients that have fewer barriers to employment could also play a role in improving placement rates and earnings. The agency’s exit data include information about exiters’ education levels and whether or not they have been incarcerated. Figure 6 shows exiters by broad education category.
The share of exiters with no college experience has shrunk slightly since 2009. Customers with high levels of formal education remain much the same, and those in the middle, having some college, an associate’s degree or other post-secondary credential slightly expanded. This changing composition of skills may suggest that EC is serving fewer customers on the lower rungs of the skill ladder; however this does not appear to be the whole story.
While the composition of the exiting population maybe changing, EC is actually exiting higher numbers of clients at all education levels. Figure 7 shows the number of exiters by level of education.
In the first full year of the demand model, Employment Connection served many more people at all educational levels. When compared with 2009, there were well over twice as many exiters with no formal education beyond high school, well over three times as many with some post-secondary education, including associates degrees and similar credentials, and well over twice as many with at least a bachelors degree.
Placement rates increased across all education categories. Figure 8 shows the percent of individuals at each level of education that entered employment by year.
EC supports formerly incarcerated citizens’ assistance programing through a contract with Towards Employment, a Cleveland nonprofit that provides job placement services and connects job seekers to training. The overall share of exiters that had been incarcerated shrank in 2011. The number of exiters identifying as being formerly incarcerated also declined after peaking in 2010. However, placement rates for this population have increased more than 188 percent, growing from 33.1 percent in 2009 to 95.5 percent in 2012. Table 8 shows these changes.
Workers who are over the age of 40 often have additional barriers to finding employment and are protected from workplace discrimination by the Age Discrimination in Employment Act (ADEA). Older workers are exiting Employment Connection programs at similar rates after the demand-facing transition at EC as they were before. Placement rates have improved but not as sharply as they have for other demographic categories. Table 9 shows these changes.
Other demographic trends are emerging at Employment Connection. More African Americans are exiting EC programs. Table 10 shows these changes.
African Americans are exiting in higher numbers and represent a larger share of total exiters. Placement rates among African American exiters increased, by 67 percent between 2009 and 2011, with more than 80 percent of exiters entering employment in both 2011 and 2012.
Women exiters have increased in number and as a share of the total exiting population. More women are exiting the EC program and higher numbers of women are exiting into employment. Table 11 shows the number of women served, the share of total exiters that are women and the placement rate of exiting women.
By many indicators, Area 3 is doing more with less funding. Placements are up and wages have increased even with little change to the regional economy. Under the demand-facing model the return on the public investment into the workforce system tripled. Table 12 shows an estimated return on investment for each year.
Overall, the ROI increased since the agency transitioned to the demand-facing model. For every Workforce Investment Act dollar invested in Area 3 there is an estimated $2.78 return to public coffers. The number of months to pay back the public investment has shrunk since fiscal year 2007, as has the cost per placement. The agency is functioning with greater efficiency.
Even though placements and exits have greatly increased, Employment Connection is seeing a decline in the number of individuals served. Figure 9 shows the total number served by program year according to WIA annual reporting.
In 2009, 8,361 individuals accessed services through the agency. By 2011 that number fell by more than 1,800 customers. The 2009 service numbers are a high-water mark. In 2008-09, during the height of the recession, the federal stimulus sent $13.8 million in additional funding to Area 3. This nearly doubled the number of youth accessing services. There were large numbers of jobseekers, few jobs, and increased funding to help adults, and particularly youth, find jobs. The 2011 service numbers are on par with 2008 levels, even though WIA funding has declined. Quarterly data suggests that the agency is maintaining service levels around 6,000. Although in the first quarter of program year 2011, the number of participants sharply fell to 3,405.
The model has improved overall efficiency, and more customers are reported as exiting the program and entering employment, but the declining WIA budget appears to be impacting the total number being served, so even as more people are exiting and outcomes are improving, the pathway to services is narrowing. Improved outcomes are crucial, but it is also important to maintain high service levels. Reversing cuts at the federal level and increasing state funding to innovative service approaches could give Area 3 the fuel to bring more customers through the agency and into training and employment. Now that the agency is doing such a good job of placing clients, it is well positioned to serve more of the many jobless Ohioans.
Conclusion and recommendations
Employment Connection’s outcomes quickly improved after implementing the demand model of service delivery. More clients are exiting the program into jobs. Earnings for EC exiters are increasing and gaining on the average regional wage even without significant improvement in the Cleveland economy and even though the agency has been exiting a higher number of clients with less education. The demand-facing model with its focus on building employer relationships appears to be a good fit for Area 3 customers and the regional economy.
The role of the business service team and the reskilling of its internal workforce played a large role in changing employer perceptions of the agency and its customers. The rebalancing of training to support short-term training also works for a client population that skews older, and are more likely to have previous work experience. The most important lesson from the Employment Connection transition is not necessarily the service delivery model; rather, it’s the idea that the service delivery model must be closely tailored to the agency’s clients, local employers, and the regional economy.
Employment Connection’s leadership says that the model works for the current labor market but the goal is maintaining flexibility to respond to changes when they happen. If the Cleveland economy returns to low levels of unemployment, with high levels of job growth in emerging industries, then the agency must respond by shifting resources to support longer-term training. If the client population starts to skew toward younger workers with no work experience, then the training policy should reflect the need for bridge to work programs.
One of Ohio’s economic strengths is that it is incredibly diverse. The state is home to urban areas, each with distinct labor traditions and to rural regions that range from agricultural economies to extraction economies in Appalachian Ohio. It should go without saying that what works in Cleveland will not necessarily work in other areas of the state. The process of building relationships with employers will be shaded by the local employer composition, but also by regional and local cultural norms. Employer outreach in Cleveland will be very different from outreach in Portsmouth, but it should be happening in every WIA Area.
The Workforce Investment Act mandates local control so the system can be tailored to local conditions. The local Workforce Investment Board is responsible for shaping local policy and is responsible for matching the delivery system to the economic realities of their area. Evaluation and change are not easy to tackle and if areas do not have relationships with workforce stakeholders, including employers, labor, and advocates, they may not have enough information to make good decisions about service delivery.
The state could play a larger role in helping local areas bridge this relationship gap. Sector partnerships function at the regional level and bring together workforce stakeholders, including employers, labor, training providers, the government, researchers, and community organizations to assess and address the workforce needs of specific industries. Sector partnerships have been linked to increases in employer productivity, and to decreases in client poverty. More than 20 states have implemented statewide strategies. Pennsylvania has supported 76 partnerships, serving over 100,000 workers, and increasing wages and job retention rates.
In Ohio, the Cincinnati region’s sector partnerships in healthcare, advanced manufacturing, and construction are driven by more than 50 employers and have trained more than 6,000 jobseekers, as well as incumbent workers, for in-demand jobs since 2008. In northwest Ohio, the Skills for Life Marine Trades Training Initiative has been nationally recognized by the US Department of Labor and has leveraged more than a million dollars in private investment to train workers to use cutting edge technology and equipment.
Sector partnerships can create the regional infrastructure needed to support the deep relationship building WIBs need to build closely tailored service delivery systems. The state should use a small component of its workforce training funds, enlarged now through casino revenues, to support regional sector partnerships. Unlike training vouchers, which benefit individual employers and are often geared toward meeting immediate, short-term training goals, sector partnerships would support regional sector-wide productivity. Such funding would provide competitive grants, to regional consortia for the development and implementation of employer and worker responsive curriculum and training. Awarding criteria should prioritize partnerships that bring together a diversity of stakeholders, such as employers, training providers, workforce and community advocates, and labor and worker representatives, and partnerships that focus on building career pathways for moving low-skilled workers into higher skilled jobs in, in-demand occupations.
The Employment Connection transition also highlights the importance of funding. Additional funding, from the federal stimulus, was needed to support the consultation and evaluation of the program, and was used to bring on staff with new skills. Funding cuts lead to fewer overall staff, and even though the agency is functioning with greater efficiency, increasing placements and the number exiting, EC is serving fewer clients. With adequate funding, the agency could expand its service capacity, increasing the number served and maintaining both individual training accounts and on-the-job training. The needs of clients and employers should drive decisions about which programs are emphasized; those choices should not be dictated by available funding streams. Efficiency is the essential stewardship of the public dollar, but it is not a replacement for adequate funding. Federal and state policymakers should take a balanced approach to public budgeting, restore revenue, and avoid further restrictions to this important set of services.
AppendicesA — Employment Connection WIA common measure reporting, program year 2011 B — Individual training account policy, November 19, 2010 C — On-the-job training program, policy and procedures, August 9, 2011 D — Estimated return on investment, fiscal years 2007-2011
We are grateful to the Joyce Foundation for generous funding for this and other research on adult worker training. The author thanks Employment Connection leadership and staff for their assistance with this report, notably David Megenhardt, Anthony Fluellen, Amy Emery, Candace James, and Frank Brickner. David Megenhardt, the United Labor Agency’s executive director, is a founding member of Policy Matters Ohio’s board of directors. Policy Matters intern Marcy Jagdeo provided research assistance for this report. As always, any errors or omissions are the responsibility of the author.
 The city and county employ 25 staff at Employment Connection. Towards Employment, a non-profit employment agency manages the Area’s returning citizens (ex-offender) work, Mature Services supports some staff as front desk, customer service staff, and in the Temporary Assistance for Needy Families program.
 Employment Connection maintains an orientation program and a career advancement skills workshop to teach many of these skills to clients. Resumes are now immediately written with the help of staff if necessary, and loaded onto the agency’s database so that the individual can be matched with potential employers.
 Increase measured from 2009 to 2011, based on EC-provided exiter data.
 Author interviews with David Megenhardt, Anthony Fullen, and Frank Brickner.
 Author interviews with David Megenhardt.
 The increased need for community college assistance is not surprising. In 2009, Ohio cut funding to the primary source of need-based state aid, the Ohio College Opportunity Grant (OCOG). The state also imposed a “Pell-first” policy, which, in effect, cut OCOG funding eligibility for students at community colleges and 2-year public branch campuses. Students struggling to complete degrees and credentials at these institutions clearly need more assistance, but job connected training must be the priority in the workforce system, as WIA dollars are also shrinking.
 Unemployment was very low in the late 1990s and unlike today’s so-called skill gap; wages actually rose in response to employer demand for greater workforce skills. See, State of Working Ohio, The Middle Falls, Policy Matters Ohio, available at http://www.policymattersohio.org/the-middle-falls.
 Ohio Department of Job and Family Services, CES data, June 2009-February 2013. February figures preliminary at time of publication.
 Ohio Department of Job and Family Services, LAUS data for February 2013.
 See, for example, the article by Motoko Rich, “Factory Jobs Return, but Employers Find Skill Shortage,” New York Times, July 1, 2010.
 To not meet a measure the agency must fail to reach 80 percent of the negotiated standard. See Table 1 for details.
 The number of people served and placed in the annual WIA reports differs from the numbers generated from EC outcome data made available to the author, which is used later in this report. See Figure 1 as an example. The outcomes data includes everyone that exited to employment, and includes placements of individuals that used services to get a better job, those individuals who entered the program employed, would not be included in the WIA outcomes. The outcomes data captures a much broader picture of the agency.
 Several factors can contribute to a perceived skill gap, including offering too low a wage for the skills sought, electronic and key word search driven application processes, loss of HR function, and geography. See “Hollowing Out: Job loss, job growth and skills for the future,” Policy Matters Ohio available at www.policymattersohio.org/hollowing-out-dec2011.
 Leadership that advocated for and guided the agency’s development included Judith Weyburne, Frank Brickner, and Larry Benders. See “Team NEO: New Economic Development Practices,” by Jane Day at IBMag.com, August 2011.
 City of Cleveland/Cuyahoga County Workforce Investment Board-Area 3, Position Paper, August 2011.
 The Cleveland Clinic prioritizes EC referrals and uses EC as its primary placement agency for entry-level workers.
 House Bill 1 and Senate Bill 2, recently passed by the 130th General Assembly now requires all One-Stops to be rebranded as OhioMeansJobs sites. The bills also require One-Stops to use the OhioMeansJobs.com site and dashboard for all placement and labor exchange functions. The bills were amended to clarify that One-Stops may continue to use and receive funding for case management systems, like EC’s, that have additional functions not duplicated by OhioMeansJobs.com. It is important that the local innovations, like the EC listing and client management system, are not unintentionally quashed in the state’s push for uniformity and higher system wide performance.
 Sources: Ohio Department of Job and Family Services, WIA Annual Reports, available at: http://jfs.ohio.gov/owd/wia/WIA-Performance-Resources.stm
 See “WIA common measures at a glance,” retrieved at wdr.doleta.gov/directives/attach/TEGL17-05_AttachA.pdf.
 See “Employment Connection WIA outcomes matrix,” provided to author by David Megenhardt, Appendix A. See also, Table 3, p. 7 for an example of WIA outcome lag.
 By sorting by exit year, the outcomes of those who left the program in 2009 could be compared to the outcomes of those that left after the transition in 2010. Based on the data provided, 2009 reflects the outcomes of people who entered and exited in 2009, 2010 reflects the outcomes of people who exited in 2010, but who could have entered the program in 2009 or 2010, 2011 data represents those that exited in 2011, but entered in 2009, 2010, or 2011. This means that there may be additional exiters in 2009, people that entered the program prior to 2009 that may not appear in this data. This methodology is different than EC uses internally, which is to sort exiters based on when they entered the program. Sorting based on entry year means that 2009 outcomes would include outcomes for those entering in 2009 but exiting in 2009, 2010, 2011, and 2012. Sorting by entering data would pull outcomes from various years into the 2009 totals. For that reason we chose to isolate outcomes by exit year.
 The outcomes data used in Figure 1 includes everyone that exited to employment, and includes placements of individuals that used services to get a better job, those individuals who entered the program employed, would not be included in the WIA outcomes. The outcomes data captures a much broader picture of the agency.
 Comparing 2009 to 2011 could be considered comparing the agency’s blooper reel with their highlight reel. The agency’s WIA outcomes for program years 2008 and 2007 were relatively mediocre, exceeding in six measures in 2008 and five in 2007. If all Ohio Areas were ranked according to the number of measures exceeded, less the number not met, Area 3 would have ranked fifth out of 20 in 2007, sixth in 2008, last in 2009, and would have been at the top of the ranking in program years 2010 and 2011.
 See, JobWatch for monthly analysis of Ohio’s job recovery, available at http://www.policymattersohio.org/category/research-policy/work-wages/sub-topics-work-wages/jobwatch.
 Infra at note 29, discussing area labor force data.
 CES data for Cleveland MSA, seasonally adjusted total nonfarm jobs, available at http://ohiolmi.com/asp/CES/CES_GET.asp. In June 2009, the MSA posted 997,100 seasonally adjusted nonfarm jobs. In November 2012, that number had grown to 998,900. At time of writing, November 2012 data was preliminary.
 It is not uncommon for the unemployment rate and the jobs count to present conflicting data. The data points come from two separate surveys. In this local example, the actual areas being measured by each survey is also different. The unemployment rate covers Area 3. The jobs number covers the Cleveland MSA. Often when these data conflict the unemployment rate is being affected not by job growth but by a change in the number of individuals included in the labor market. The Civilian Labor Force Estimate annual averages for Area 3 show that the labor force has shrunk by 12,800 since 2007.
 Day, Jane, “Team NEO: New Economic Development Practices,” IBMag.com, August 2011.
 Position paper, supra at note 18.
 Customized and on-the-job programing was allotted nearly 30 percent of the all funds training budget in FY2011.
 See Community Research Partners, “Local WIA Policy in Ohio,” September 2011, retrieved at www.communityresearchpartners.org/uploads/publications//WIA%20Final%209-12-11.pdf. The report contains a detailed cataloguing of state and local WIA policy in Ohio. See also Randall W. Eberts, “Individual Training Accounts Provided under the US Workforce Investment Act,” W.E. Upjohn Institute for Employment Research, December 2010.
 Id. at p.13.
 WIA Title 1.B.134.
 Supra, note 33 at p. 12.
 Individual Training Account (ITA) policy, Employment Connection, Approved 11/19/2010, at p. 3, Appendix B.
 Author interview with Robin Rice, Office of Workforce Development, Ohio Department of Job and Family Services.
 Mathematica Policy Research, Larger Training Awards and Counseling Improve Cost Effectiveness, Long Term Outcomes of Individual Training Accounts, April 2012 available at http://www.mathematica-mpr.com/publications/PDFs/Labor/ITA%20Fact%20Sheet_042512.pdf. The study compared the long-term earning and savings outcomes of structured Individual Training Accounts (customized, higher awards and required counseling), to guided-choice ITAs (fixed, lower awards and time caps, and required counseling), and maximum choice ITAs (with a fixed, lower award but only voluntary counseling). The study found that structured ITAs could generate an additional $41,000 in earnings through retirement age for participants, or about $500 extra per quarter than the other models. Structured ITAs also lead to savings, as higher earnings generated higher tax revenue offsetting program costs.
 Employment Connection, On-the-Job Training program policy and procedures, at Appendix C.
 See “Outcomes matrix,” Appendix A.
 See, National Skills Coalition, State Sector Strategies Coming of Age: Implications for State Workforce Policymakers, January 2013.