Who Leaves, Where To, And Why Worry?: Employee Mobility, Employee Entrepreneurship, And Effects On Source Firm Performance

1
Who Leaves, Where To, And Why Worry? Employee
Mobility, Employee Entrepreneurship, And Effects
On Source Firm Performance

• Benjamin A. Campbell
• Ohio State University
• Martin Ganco
• University of Illinois
• April M. Franco
• University of Toronto
• Rajshree Agarwal
• University of Illinois

2009 Joint U.S.-Canadian Census Research Data
Center Conference October 5, 2009
2
The small print

• The research in this paper was conducted while
  Ben Campbell and Martin Ganco had Special Sworn
  Status as researchers of the U.S. Census Bureau
  at the Chicago Census Research Data Center.
   Research results and conclusions expressed are
  those of the authors and do not necessarily
  reflect the views of the Census Bureau.  This
  research has been screened to insure that no
  confidential data are revealed.

3
Motivation

• Where does productivity come from?
• Recent research points to the importance of human
  capital and human assets
• Entrepreneurs previous work experience helps to
  determine the success of their ventures
• But several questions are yet unanswered
• Which individuals are most likely to leave a
  firm?
• Which individuals are most likely to go to a
  spin-out?
• What is the impact of such movement on the
  incumbent firm?

4
Employee mobility in knowledge intensive
industries
High-Technology Manufacturing
Human Capital Intensive Services

• Klepper and Sleeper, 2004
• Franco and Filson, 2006
• Agarwal, Ganco and Ziedonis, 2008
• Agarwal, Echambadi, Franco, and Sarkar, 2004
• Elfenbein et al., 2008
• Klepper and Thompson, 2009

• Somaya et al., 2007
• Wezel et al., 2006
• Phillips, 2002
• Groysberg et al., 2007

5
Firm organization and employee mobility in the
services sector

• Garicano and Hubbard, 2007
• Levin and Tadelis, 2005
• Rebitzer and Taylor, 2007

6
Main Ingredients

• Bargaining Power
• Relative importance of complementary assets to
  production
• Ability to transfer/recreate complementary assets
• Create New Opportunities
• Ability to reconfigure complementary assets to be
  more productive

7
Who Leaves, where do they go, and does it matter??

• H1 There is a negative relationship between
  earnings and the likelihood of employee mobility.
• H2 Conditional on mobility, individuals with
  greater earnings are more likely to join
  spin-outs than join established firms.
• H3 The adverse impact on firm performance due
  to employee mobility is greater for employee
  mobility to spin-outs than for employee mobility
  to incumbents.
• H4 The adverse impact on parent firm
  performance due to employee mobility to spin-outs
  relative to employee mobility to established
  firms increases with the earnings of the moving
  individual.

8
Context

• We test the hypotheses in the legal services
  industry
• Human capital is easily transferrable (within
  state borders).
• Overhead costs are low, and wage bill is the
  dominant cost, hence the aggregate wage bill is a
  good proxy for revenues (Gilson and Mnookin,
  1985).
• Data
• A custom extract of the Longitudinal
  Employer-Household Dynamics (LEHD) Project
  available at the Census Research Data Centers.
• The data are longitudinal spanning over 10 years
  and covering 10 large states.
• The custom extract includes all workers who have
  ever worked in the legal services industry and
  all firms that have ever reported operating in
  the legal services industry.

9
Key Variables

• Mobility to Spin-out/Mobility to Incumbent
• longitudinal records of employment history allow
  us to track employee mobility and employee
  entrepreneurship.
• Employee Earnings
• Earnings include all forms of taxable
  compensation that are received in the calendar
  year, including salary, bonuses and other
  reported income.
• Firm Performance By summing the earnings of the
  universe of employees inside the firm, we
  re-construct the total revenues earned by the
  firm. We divide by employment to capture the
  revenue generated per employee (in 10,000s).
• Firm Mobility Measures. We measure mobility in
  three ways
• Number of employees leaving
• Aggregate Payroll of employees leaving
• Number of employees leaving in different payroll
  classes
• 0-100K
• 100K-300K
• 300K-5000K

Individual
Firm
10
Universe

• For the individual-level analysis
• a random 25 sample of the data
• individuals who earn more than 25K annually
• Individuals who are employed at a firm of more
  than 5 people
• Individuals that are employed at a firm that does
  not exit the data that year or in the subsequent
  year
• For the firm-level analysis
• firms with more than 5 employees
• firms do not exit in the current or subsequent
  year
• firms that have revenue per worker of between
  10K and 1M
• firms that do not lose greater than 20 workers in
  any payroll class to an established form or to a
  startup in a given year

11
Empirical Strategy

• Stage 1 (Who Leaves? To Go Where?)
• First, identify the individual characteristics
  that are related to employee mobility in general.
• Second, identify the individual characteristics
  that are related to mobility to spin-out
  conditional on mobility.
• We estimate a series of linear probability model
  with firm-year fixed effects and robust standard
  errors.
• Conditional logit is computationally infeasible
  with our sample size.
• Out-of-sample predictions are extremely rare
  indicating that the linear probably model
  performs acceptably.
• Stage 2 (And Does it Matter?)
• Firm performance is a function of the intensity
  of different types of worker mobility and firm
  characteristics.
• We estimate a series of linear regression models
  with firm fixed effects and year dummies.

12
Who leaves, to go where?
?
H1 supported
H2 supported
13
And Does it Matter?Mobility to spin-out vs.
Mobility to incumbent
H3 supported
The average 85-employee firm faces a 22,865 loss
when an employee moves to a spin-out!
14
And Does it Matter?Value appropriation and
mobility to spin-out
If the employee earns between 100,000 and
300,000 and moves to a spin-out, the firm
faces a 193,000 loss in revenues!
H4 supported
If the employee earns between 300,000 and
1,000,000 and moves to a spin-out, the firm
faces a 1,000,000 loss in revenues!
15
Summary

• Workers with higher earnings are less likely to
  move (H1)
• If workers with higher earnings do move, they are
  more likely to go to startups (H2).
• Mobility to spin-out has a larger adverse impact
  on parent performance than mobility to incumbent
  (H3).
• The adverse impact on firm performance of
  mobility to spin-out increases with earnings of
  the mover (H4).

16
Contributions

• We study the implications of knowledge transfer
  mechanisms where knowledge is a rival good
• We provide a direct comparison between mobility
  to incumbent firms and mobility to spin-outs,
  both at the individual and at the parent firm
  level

17
Follow-on Projects

• Working on
• Impact of Teams
• Does it matter if a superstar leaves with a team
  or as an individual?
• Does a superstar hurt the firm more if she goes
  to a start-up?
• Compensation practices and spin-out/incumbent
  performance

18
(No Transcript)
19
(No Transcript)
20
Descriptive Statistics-Individual
21
Descriptive Statistics-Firm