ta Quality Institute: Perkins 101, Harvesting Real Time Employment ...

Harvesting Real Time Employment Data 2015 Data Quality Institute November 5, 2015

Objectives Scott Murakami, Director of Workforce Development, Office of the Vice President for Community Colleges – – – –

What is real time data? What are we doing with it? Why are we doing it? How are we using it?

Scott Wheeler, Director of System Performance, State of Washington 2

Role at the University of Hawaii Community Colleges • Provide campuses and partners with baseline economic/workforce data. The data is used for planning and serves as a starting point for discussion with business and industry to align education and training programs to industry needs. • Sector analysis and ensuring that our analysis is consistent with the economic direction of the state and counties of Hawaii. • Liaison between the UHCC and the Hawaii Chamber of Commerce and business and industry. • Liaison between the UHCC and the public workforce system 3

Basic Rules on all LMI • LMI both real time and forecast (structural) are the starting point for discussion with business, industry and government agencies. • Real time LMI is not a replacement for structural LMI. It is another tool that allows us to examine the workforce from a different dimension. – Not looking for a silver bullet. We are looking to improve analytics by complementing our toolbox.

4

What is Real Time Data? • Big data tool for looking at labor market Information. • Data tool that allows us to analyze qualitative data to supplement the quantitative projection analysis based on structural data on labor market activity in Hawaii. • Aggregates, scrubs, normalizes, and manages job posting data to be used in analysis that helps users identifies:

– Who is hiring; What employers are hiring – How many; Number of job openings (unique and duplicated) – What are employers hiring for; Common in-demand skills 5

How does it work? • Spiders online job posting • Using Natural Language Processing, identifies similarities in posting and cleans (scrubs) heterogeneous qualitative data. • Normalizes the data into fields such as, job titles, skills, credentials, educational attainment, and years of experience. • Data engine that allows the user to search the clean, normalized data and create customized reports by geographic region and timeframe. 6

Hawaii’s Use of Real Time LMI • Real Time LMI is supplementing our planning process – Economic Development Planning

• Cluster economic development strategies

– Workforce Planning

• Industry-based sector strategies

– Institutional Planning

• Program planning • New program development • Short-term program planning

• Accountability: Forthcoming – Workforce mapping project

• Hierarchy of occupations within industry sub-sectors that are mapped to both local skills, compensation and educational attainment

– Perkins Indicators

7

How Are We Using It? • Assessment of employers

– Frequency and intensity of posting

• How many unique postings vs. the intensity of posting

– Market demand for occupations vs. the desire to place these positions

• Local skills assessment

– Supplements campus program advisory boards with data that is validated by the larger industry

• Geography of the postings

– Is there consistency with the cluster economic activity? – What are other logistical issues the workforce faces? 8

Hawaii’s Sources of Real Time LMI • Economic Modeling Specialists International (EMSI) - job posting analytics • Evaluating burning glass – labor insight/jobs for procurement

9

Why the Addition of Real Time LMI? — Macro Perspective Ongoing Shift in Higher Education from completion to job placement • The Completion Agenda to the White House Ready to Work Initiative • Indicators – – – – –

Gainful employment ARRA grants Checklist for job-driven training TAACCCT grants outcome measures College scorecard

• “What can we do to cause completion and graduation?” vs. “What are the factors that contribute to job placement for our graduates?” – Shift from causation to correlation – Move from projection to prediction

• Clean qualitative data can supplement the forecast data in our analytics 10