Supplementary Analyses: Member Turnover and Dwell Time Denis O. Gray Olena Leonchuk Lindsey McGowen IUCRC Evaluation Project Slide 1
Goals Obtain a better understanding of the factors that might be affecting turnover of IUCRC members Aggregate Statistics Obtain a better understanding of the factors that affect dwell time or the continuous maintenance of IUCRC membership Individual Member Statistics IUCRC Evaluation Project Slide 2
89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 Members Average Membership Turn Over NEW CENTERS EXCLUDED Lines crossing = usually means recession! 6 5 4 3 Members Added this FY 2012 Leaving rate: 20.3% Small = 24.83% Federal = 20.63% Large = 17.87% State = 7.14% Other = 6.67% Members Left this FY 2 1 0 Fiscal Year January, 2013 *Newly funded members are not counted as Members Added FY2011-2012 NSF-I/UCRC Center Structure Database Slide 3
Member Turnover Rate 30.00% 25.00% 20.00% 15.00% Turnover Percent Project-Based Consortial 10.00% 5.00% 0.00% 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Turnover % = Members terminated in year X+1 / Total members in year X NSF-I/UCRC Center Structure Database January, 2013 Slide 4
Member Turnover Rate New Normal 30.00% 25.00% 20.00% 15.00% Turnover Percent Project-Based Consortial 10.00% 5.00% 0.00% 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Turnover % = Members terminated in year X+1 / Total members in year X NSF-I/UCRC Center Structure Database January, 2013 Slide 5
Member Turnover Rate New Normal 30.00% 25.00% 20.00% 15.00% Turnover Percent Project-Based Consortial 10.00% 5.00% 0.00% January, 2013 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Working Hypotheses: - National recessions - 1990s vs. 2000s - Mode of operation: project vs. consortial - Small/SBIR Turnover % = Members terminated in year X+1 / Total members in year X NSF-I/UCRC Center Structure Database Slide 6
Dwell Time Analyses January, 2013 NSF-I/UCRC Center Structure Database Slide 7
Method Sampling & Time Period that started between 1992 and 2001 and that had at least 9 years of NSF support (collection of data) Members of these centers between 1992 and 2011 Procedure Combine data from archival membership database (1992-2006) and the latest membership data from 2007-2011 Assign an individual numeric code for each member Calculate consecutive years of membership Classify members into 7 categories: 1. Large Industry (>500 employees) 2. Small Industry (<500 employees) 3. Government (non-us) 4. Government (US Federal) 5. Government (US State/Local) 6. Non-Profit 7. Other Organization [unidentified organizations] NSF I/UCRC Membership Dwell February 13 8 Time
Results Descriptive Total 2121 individual memberships were identified from which: Large Industry 55% Small Industry 14% Government (U.S. Federal) 10% Government (U.S. State/Local) 2% Government (non-us) 1% Non-Profit 4% Other Organization 14 Average of the consecutive years companies maintained their membership at I/UCRCs is 3.74 NSF I/UCRC Membership Dwell February 13 9 Time
Percentage of organizations Member dwell time Percent of organizations maintaining an I/UCRC membership for X number of years (Mean dwell time = 3.8) 120% 100% 80% 60% 40% 20% 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 N of years as a member January, 2013 NSF-I/UCRC Center Structure Database Slide 10
Dwell Time (Every Other Year) by Member Starting Year(corrected for 5 in 1995) 80 Membership Dwell Time of groups of companies from beginning years between 1993-2008 (every other year) 70 60 50 40 30 3 5 7 >7 20 10 0 1993 1995 1997 2000 2002 2004 2006 2008 IUCRC Evaluation Project Slide 11
N of Members Members by percentage of possible dwell time 600 500 400 300 200 100 0 Dwell time % (Mean = 32.19%) Dwell time % = Total N of years as a member / Total possible N of years as a member NSF-I/UCRC Center Structure Database January, 2013 Slide 12
Limitations Missing years in membership data imputed data Maddening variations in member names to reconcile Sample issues: exclude centers with less than 9 years data Different start and end years of the NSF funding that determines data collection by the NC State evaluation team Disaggregated analyses (by type of member and center) still remains! February 13 NSF I/UCRC Membership Dwell 13 Time
Conclusions: Turnover Member turnover is becoming an increasingly big challenge for IUCRCs Longitudinal data on membership turnover shows increased percentage of members leaving in recent years Possible explanations are: Internal: Increasing percentage of Small Businesses leaving after receiving the SBIR support and not being able to sustain membership independently More opting for a project-based mode of operation External: Globalization or development of communication technologies create more alternatives for companies. Periodic recessions making memberships unstable IUCRC Evaluation Project Slide 14
Conclusions: Dwell Rate Multiple indicator strategy for dwell rate Percentage maintaining membership steadily declines over time Average Dwell Rate = 3.8 years Average member stays 32% of possible years No discernible pattern by year members started membership (1993-2008) More detailed research needs to be done to identify increased turnover rate of the members using more precise classification of the members and centers IUCRC Evaluation Project Slide 15