labor market advantages of organizational status: a study ...€¦ · large u.s. law firms...
TRANSCRIPT
Labor market advantages of organizational status:
A study of lateral partner hiring by large U.S. law firms.
Christopher I. Rider
Goizueta Business School
Emory University
David Tan
Foster School of Business
University of Washington
January 31, 2014
* Order of authorship is alphabetical. We appreciate helpful comments from Joe Broschak, Diane Burton, Seth
Carnahan, Olivier Chatain, Henrich Greve, Giacomo Negro, Peter Thompson, reviewers, and audiences at
Boston University, the University of Chicago, Copenhagen Business School, Duke University, the University
of Michigan, Washington Univ. in St. Louis, the 7th
Institutions & Innovation Conference, the 3rd
People &
Organizations Conference, and the 14th Meeting of Organizational Ecologists. Financial support from Emory
University’s Goizueta Business School and the Law School Admission Council is gratefully acknowledged.
Labor market advantages of organizational status:
A study of lateral partner hiring by large U.S. law firms.
Abstract
Prior research demonstrates product market advantages of organizational status but largely neglects factor
market advantages. We propose that status is advantageous in labor markets because individuals generally
consider employer status a non-pecuniary employment benefit. Dyadic analyses of lateral partner hiring by
large U.S. law firms demonstrate two status-based advantages in employee hiring and retention. First, high
status firms are more likely than low status ones to hire an employee from a more profitable competitor.
Second, high status firms are most likely to lose an employee to a lower status competitor when the
competitor is – atypically – more profitable. We discuss implications of these findings for individual and
organizational status attainment and for the stability of industry status hierarchies.
1
Fundamental to the status-based model of market competition is the notion that organizational
status is a basis for product market advantage (Podolny, 1993; 2005). When product or service quality is
difficult to verify ex ante – frequently for cultural goods (Benjamin and Podolny, 1999; Bielby and
Bielby, 1999) or professional services (Podolny, 1993; 1994; 2001; Chung, Singh, and Lee, 2000; Jensen,
2006; Jensen and Roy, 2008; Uzzi and Lancaster, 2004) – market actors generally infer quality from the
status of the organization offering it for sale. Consequently, relative to low status competitors, high status
organizations enjoy lower marketing costs, higher prices, and greater sales, resulting in a typically
positive correlation between organizational status and profitability (Podolny, 1993; Podolny, Stuart, and
Hannan, 1996; Benjamin and Podolny, 1999; Podolny and Scott-Morton, 1999).
Prior research does not similarly establish status-based advantages in factor markets, despite
interest in such advantages. Most prominently, Podolny (2001: 43) speculates that hiring and retention
costs for equivalent labor are decreasing with status because individuals generally view high employer
status as a non-pecuniary employment benefit. Low status organizations are, consequently, constrained to
hiring from highly uncertain market segments (i.e., inexperienced candidates). Recently, Bidwell, et al.
(2013) argue that, relative to low status competitors, high status employers can hire more capable junior
employees at equivalent cost because individuals expect future career opportunities to be increasing with
employer status. But, despite extensive recent research on hiring competitors’ employees (e.g., Rao and
Drazin, 2002; Broschak, 2004; Wezel, Cattani, and Pennings, 2006; Groysberg, Nanda, and Lee, 2008;
Dokko and Rosenkopf, 2010), the advantages of status in labor markets remain largely speculative.
In this study, we examine whether or not organizational status aids in employee hiring or
retention. We draw on the economic concept of equalizing differences (Rosen, 1986), which suggests
that individuals trade-off pecuniary and non-pecuniary benefits in employment alternatives (e.g., Stern,
2004). We specifically consider how, by providing a basis for non-pecuniary benefits, organizational
status aids in hiring competitors’ employees and in retaining employees valued by competitors.
Assumptions of individual and organizational preferences inform baseline predictions about which
organizations within an industry are most likely to hire each other’s employees. We theorize a central
2
tendency of organizations to hire employees from competitors of similar status and profitability. But, we
also argue that deviations from these tendencies entail individual trade-offs between pecuniary and non-
pecuniary employment benefits that favor high status organizations over lower status ones.
We situate our inquiry in U.S. legal services, an industry characterized by increased hiring from
competitors over the past decade (see Figure 1). Specifically, we analyze data on thousands of lateral
partner hires among the largest U.S. law firms between 2000 and 2009. We find that organizations do
tend to hire employees from competitors of similar status and profitability, but two labor market
advantages of organizational status are clear. First, high status firms are more likely to hire a partner from
a more profitable competitor than are low status firms. Second, high status firms are most likely to lose a
partner to a lower status competitor when the competitor is atypically profitable, given its status. Extant
theory suggests that the hiring advantage enhances profitability and that the retention advantage maintains
status. We attribute both advantages to employer status as a valuable non-pecuniary employment benefit.
-----------------------------------------
Insert Figure 1 About Here
-----------------------------------------
We discuss the implications of our study for intraprofessional status mobility (i.e., individual
transitions to employers of different status), for our understanding of the commonly observed positive
correlation between organizational profitability and organizational status, and for the stability of industry
status hierarchies. These considerations motivate a research agenda on organizational status attainment
that parallels the extensive literature on individual status attainment (see Lin, 1999 for a review).
Theoretical Development
The status-based model of market competition introduced the idea that organizational status
signals product or service quality in markets where quality is either difficult or costly for consumers to
verify ex ante (Podolny, 1993). Ex ante, consumers generally expect offering quality to be increasing
with producer status. Ex post, high status producers also receive greater recognition and rewards for work
of equivalent quality than low status actors receive because status enhances visibility and positively biases
3
the allocation of credit (Merton, 1968). In the context of market competition, these tendencies imply that
while nothing in principle prevents low status organizations from investing in high quality, high status
organizations realize greater returns on such investments (Podolny, 1993; Benjamin and Podolny, 1999).
High status producers consequently realize higher prices for offerings of equivalent quality.
These effects are especially pronounced in professional services, such as investment banking (Podolny,
1993; 1994; 2001; Chung, et al., 2000), accounting (Jensen, 2006; Jensen and Roy, 2008), and corporate
law (Uzzi and Lancaster, 2004) as well as in markets for cultural goods in which quality is also difficult to
verify ex ante. For example, wine of equal quality tends to be priced higher for high status producers than
for low status ones (Benjamin and Podolny, 1999). Furthermore, wineries that hire winemakers from
prominent competitors command higher prices, post-hire, for wines of a given quality – even wines made
by prior winemakers (Roberts, Khaire, and Rider, 2011).
The advantages of status are not limited to revenue. High status organizations also incur lower
costs of marketing equivalent quality offerings because exchange partners are more willing to enter
exchange relationships the greater the organization’s status is (Podolny, 1993). For example, low status
venture capital firms must offer startups better financing terms than high status competitors in order to
secure equity investments (Hsu, 2004). Both revenue and cost advantages are, therefore, implicated in the
widely-observed positive correlation between organizational profitability and status (see Podolny, 2005).
Given these product market advantages of status, profit-seeking organizations often invest
resources in attaining higher status and in maintaining status, primarily by adopting principles of
exclusivity in inter-organizational relations (Goode, 1978; Podolny, 1993). Generally, affiliating with
lower status organizations diminishes organizational status while affiliating with higher status ones
enhances status (Podolny and Phillips, 1996). For example, status-anxious organizations quickly
disassociated themselves from discredited auditor Arthur Andersen following the Enron scandal (Jensen,
2006) and wineries were more likely to publicize a winemaker’s prior employer the greater the prior
employer’s status (Roberts and Khaire, 2009).
4
Labor markets enable profit-seeking organizations to strategically invest in status. First,
individual transitions between employers transfer status by implicitly affiliating one’s current and prior
employers (Baty, et al., 1971; Sørensen, 1999). Second, especially when quality is difficult to assess ex
ante, employees’ prior employment affiliations are effective product market signals of organizational
quality (Burton, Sorensen, and Beckman, 2002; Higgins and Gulati, 2003; 2006). Third, organizations
often transfer status by hiring competitors’ employees (Rao and Drazin, 2002; Wezel, et al., 2006;
Somaya, Williamson, and Lorinkova, 2008; Dokko and Rosenkopf, 2010; Bidwell and Briscoe, 2010).
If organizations hire employees to enhance status and profitability, then high status organizations
may be particularly vulnerable to losing employees to competitors but also likely to hire employees from
competitors. Below, we accordingly consider how status and profitability differentials between
organizations influence the baseline likelihood that one organization hires an employee from another one.
We then consider how organizational status aids in the hiring and retention of employees.
Status differentials and the likelihood of hiring. We first make explicit assumptions about
individual and organizational preferences. First, affiliating with higher status organizations generally
enhances organizational status (Podolny and Phillips, 1996) and attracts exchange partners (Burton, et al.,
2002). More specifically, prior research demonstrates that hiring individuals previously employed by
high status employers enhances market actors’ evaluations of the hiring organization. For example,
biotech executives’ prior employment affiliations to high status organizations appeal to investors and,
consequently, underwriters (Higgins and Gulati, 2003; 2006). Another study found that hiring
winemakers from prominent competitors enabled wineries to command higher prices (Roberts, et al.,
2011). Based on such findings, we assume that, all else equal, organizations prefer to hire individuals
from higher status competitors than from lower status ones.
Second, the individual desire for greater social standing is generally considered universal (Frank,
1985; Hogan and Hogan, 1991). With respect to employment, individuals typically aspire to greater
intraprofessional status and, therefore, seek work with high status organizations (Abbott, 1981; Elsbach
and Glynn, 1996; Heinz, et al., 2005). The benefits of working for a high status employer are not merely
5
psychological. Because socioeconomic rewards are also allocated in ways that favor employees of high
status organizations (Merton, 1968), individuals can also expect that future career opportunities are
enhanced by employer status (Phillips, 2001; Bidwell, et al., 2013). All else equal, then, employer status
is a valuable non-pecuniary employment benefit; individuals generally prefer to work for organizations
that are higher in status than their current employers.
These assumptions imply that, with respect to organizational status, a focal organization’s
willingness to hire an individual is generally increasing as an individual’s willingness to work for the
organization is decreasing. We expect that these countervailing preferences will tend to restrict inter-
organizational hiring to organizations of similar status because individuals generally like their status-
conferring actions to be reciprocated (Gould, 2002) and because unsuccessful hiring attempts are costly
for organizations. We, therefore, expect that the higher or lower a competitor is in status than an
organization, the less likely the competitor is to hire an employee from that organization.
Hypothesis 1: An organization is less likely to hire an employee from a competitor the greater is
the difference in status between the organization and the competitor.
Profitability differentials and the likelihood of hiring. The status differential prediction is
motivated by differences in the non-pecuniary benefit of employer status. Differences in pecuniary
benefits are, of course, also relevant. All else equal, more profitable organizations can afford to allocate
greater financial resources to pecuniary benefits than less profitable organizations can.1 For example,
empirical studies tend to find that employee compensation is increasing with employer profitability (e.g.,
Deckop, 1988; Gerhart and Milkovich, 1990; Abowd, Kramarz, and Margolis, 1999). We, therefore,
assume that pecuniary benefits are increasing with organizational profitability so that individuals
generally prefer working for organizations that are more profitable than their current employers.
Organizations, however, typically prefer hiring from more profitable competitors than from less
profitable ones because, on average, employee productivity is presumably increasing with employer
1 Of course, organizations may increase profits by allocating smaller amounts of producer surplus to employees’
wages. Our point is merely that more profitable organizations are in better positions to raise employee wages than
less profitable organizations are.
6
profitability (Sørensen and Sorenson, 2007). As in the case of status differentials, countervailing
preferences of organizations and individuals will tend to constrain lateral hiring to organizations of
similar profitability. We, therefore, expect that the higher or lower a competitor is in profitability than an
organization, the less likely the organization is to hire an employee from that competitor.
Hypothesis 2: An organization is less likely to hire an employee from a competitor the greater is
the difference in profitability between the organization and the competitor.
Two status-based, labor market advantages are implied by Hypotheses 1 and 2. First, few
competitors of high status organizations are able to offer a superior non-pecuniary benefit (i.e., employer
status). Second, if profitability and status are indeed positively correlated then few competitors of high
status organizations are profitable enough to offer superior pecuniary benefits (i.e., compensation).
Consequently, we expect that organizational status aids in hiring employees from competitors and also in
retaining employees who competitors wish to hire. We consider these implications below.
Status advantages in hiring. We first consider how organizational status facilitates the hiring of
competitors’ employees. Employee recruitment is central to an organization’s competitive advantage. As
prior work demonstrates, hiring is an effective strategy for acquiring knowledge (e.g., Almeida and
Kogut, 1999), business relationships (e.g., Somaya, et al., 2008; Dokko and Rosenkopf, 2010), and
capabilities (e.g., Rao and Drazin, 2002). Yet, hiring advantages based on organizational status have
gone largely unexplored in prior research.
We assume that organizations prefer hiring employees from more profitable competitors than
from less profitable ones because of presumed employee productivity differences that vary systematically
with organizational profitability (Sørensen and Sorenson, 2007). The theory of equalizing differences
implies that individuals considering a move to a less profitable employer will probably expect greater
non-pecuniary benefits (Rosen, 1986; Fersthman and Weiss, 1993; Stern, 2004). In such cases, non-
pecuniary benefits associated with employer status probably help high status organizations hire a more
profitable competitor’s employee. As Coleman (1990: 130) observes, ‘‘…the awarding of status to
balance unequal transactions or to make possible half-transactions appears to be the most widespread
7
functional substitute for money in social and political systems.’’ Consequently, we expect that a high
status organization is more likely to hire an employee from a more profitable competitor than a low status
organization is. Such hires are, presumably, profitability-enhancing.
Hypothesis 3: An organization is more likely to hire an employee from a more profitable
competitor if it is higher in status than if it is lower in status than the competitor.
Status advantages in retention. We now consider how status helps organizations retain employees
valued by competitors. Employee retention is critical for sustaining an organization’s competitive
advantage (e.g., Coff, 1997), as prior work demonstrates that employee departures can lead to the
dissolution of business relationships (e.g., Broschak, 2004) or entire organizations (e.g., Phillips, 2001;
Wezel, et al., 2006). Yet, status-based retention advantages have also been neglected in prior work.
We previously assumed that high status organizations’ employees are valued by competitors
because high status affiliations enhance organizational status (Podolny and Phillips, 1996; Jensen, 2006).
Low status competitors may seek to invest in status by hiring a higher status organization’s employees.
Such hires are, presumably, status-enhancing for the hiring competitor and status-diminishing for the
organization (e.g., Podolny and Phillips, 1996). How can a low status competitor facilitate such moves?
Again, the theory of equalizing differences implies that individuals considering a move to a lower
status employer will expect to be compensated for a reduction in employer status with greater pecuniary
benefits. Therefore, a competitor trying to hire an employee away from a higher status organization must
offer greater pecuniary benefits than the employee’s current employer provides. If profitability enables
organizations to make such attractive offers, then high status organizations are most at risk of losing
employees to their more profitable competitors.
Recruiting a less profitable competitor’s employees may seem a profit-diminishing strategy. But,
in the long run, such status investments could plausibly increase organization profitability by enhancing
organizational status and diminishing competitor status, thereby reducing the higher status organization’s
product market advantage. Low status organizations should, therefore, be particularly motivated to hire
employees away from high status ones. For example, although low status universities incur greater costs
8
of employing equivalent quality faculty than high status universities enhanced institutional status may
justify incurring higher costs (Meyer and Rowan, 1977; Podolny, 2001).
If status enables organizations to offer greater non-pecuniary benefits than lower-status-but-more-
profitable competitors offer, then low status organizations have weaker bargaining power with a given job
candidate than high status competitors do (Podolny, 2001; Phillips, 2001; Bidwell, et al., 2013). We
consequently expect that a high status organization is more likely to lose an employee to a lower status
competitor when the competitor is also more profitable than the high status organization. This reasoning
implies that high status organizations are most vulnerable to losing employees to competitors that are
atypically profitable, given their status (i.e., hierarchical anomalies).
Hypothesis 4: An organization is more likely to lose an employee to a lower status competitor if
the competitor is more profitable than if the competitor is less profitable than the organization.
To recap, we propose that despite individual and organizational preferences to the contrary, most
inter-organizational employment transitions are between organizations of similar status and profitability.
We also argue that by enabling organizations to offer employment with superior non-pecuniary
employment benefits, status provides two labor market advantages over competitors. First, status aids in
hiring employees of more profitable competitors. Second, although greater profitability helps
organizations hire employees from higher status competitors, the second labor market advantage of status
is that status helps organizations retain employees recruited by more profitable but lower status
competitors. Below, we outline empirical tests of these arguments.
Empirical Setting and Analysis
The context for testing these predictions is the U.S. legal services industry. We analyze hiring
events involving partners employed by the 200 highest-grossing law firms in the U.S. between 2000 and
2009 (i.e., the AM Law 200). These large, corporate-oriented, U.S. law firms are typically organized as
partnerships in which firm partners generate business, share profits, and supervise associate lawyers. Firm
status and profitability are particularly relevant for hiring because lawyers are particularly status-
9
conscious and many face large debt obligations (Phillips and Zuckerman, 2001; Sauder, 2008).
Furthermore, firm status is regularly assessed by industry surveys and so, too, is firm profitability.
Traditionally, law firms followed the “Cravath model” by growing internally, promoting
associates, developing partners, and fostering client relationships (Galanter and Palay, 1991). Associates
were promoted if partners felt that they would successfully transition from technical legal work to partner
responsibilities. The typical law school graduate joined a major law firm and hoped to become a partner
after six to eight years. Those who did would typically stay with the firm for the remainder of their
professional careers; those who did not would depart. Rarely would lawyers be hired into a partnership
from outside of the firm or dismissed after promotion to partner (Heinz, 2009).
Hiring competitors’ partners has long been a law firm growth strategy (e.g., McEvily, Jaffee, and
Tortoriello, 2012). But, until recently, such lateral hires were considered counter-normative because
lateral hiring appropriates a competitor’s investments in employees (Hillman, 2002). This norm eroded as
firm scale and competition increased and the costs of adhering to the Cravath Model’s “up or out” rule
grew larger (Sherer and Lee, 2002). Over the past decade, lateral hiring has increasingly become an
effective strategy for enhancing firm growth and performance because the work that partners perform
does not fundamentally vary across employers, but prices do vary (Uzzi and Lancaster, 2004).
Figure 1 illustrates the trend in lateral partner hiring from 2000 to 2009. In 2000, 64 percent of
partner hires by AM Law 200 firms were from an employer outside of the AM Law 200, such as in-house
corporate legal departments, government, or smaller firms. By 2009, however, the modal partner
transition was from one AM Law 200 firm to another (i.e., 59 percent of all partner hires).
Firm motivations for hiring laterally vary but can generally be characterized as status-seeking and
profit-seeking. An industry search consultant remarked in 2013, “…reasons number one, two, and three
are to buy business” (Li, 2013). One firm chairman justified hiring a partner from a more prestigious
competitor as a way to “make a mark” in the industry (Koppel, 2007). In 2006, an industry observer
described lateral hiring as “skim[ming] the cream of partners from less profitable firms” (Triedman,
2006). As for individuals, a survey of over 900 partners who moved laterally indicated that 90 percent
10
either maintained or increased their compensation but also that partners who did not move were similarly
satisfied with their compensation (Major, Lindsey, and Africa, 2012). These results imply that partners
probably view pecuniary and non-pecuniary employment benefits as compensatory. Moreover, our
assumption of a positive correlation between firm profitability and compensation is appropriate for this
setting because firm partners are typically residual claimants on firm profits (Galanter and Palay, 1991).
In sum, lateral hiring by large U.S. law firms is an excellent setting for identifying status-based labor
market advantages.
Dependent variable. We test predictions about an organization’s likelihood of hiring from a
competitor as well as its likelihood of losing an employee to a competitor. The main dependent variable in
our analysis is the likelihood that an AM Law 200 firm i hires a partner from another AM Law 200 firm j
in a given year (we use “firm i” and “firm j” throughout to denote the hiring and source organizations,
respectively, in a dyad at-risk of a lateral hire). Data on lateral hiring were obtained from Incisive Legal
Intelligence’s Lateral Partner Moves Database, which summarizes hiring information from The American
Lawyer’s annual Lateral Partner Survey of AM Law 200 firms, industry publications, firm websites, and
press releases. Between 2000 and 2009, 9,465 partners transitioned between two AM Law 200 firms.
Sample and estimation. Our unit of analysis is a dyad in which a given firm i is at risk of hiring
a partner from firm j. Our dyadic sample covers 178 unique firms and 5,270 lateral partner hires over an
8-year period from 2002 to 2009. These figures are lesser than the 200 AM Law ranked firms and the
total 9,465 lateral partner hires because we restrict the analyzed sub-sample to include lagged variables
and because firm status measures are unavailable for the lowest status firms in the AM Law 200. This
produces 162,696 dyadic, firm-firm observations; hiring occurs in 3,473 of these dyads (2.1 percent).2 For
dyads in which hiring does occur (5,270 hires from 3,473 dyads), the median and modal number of
2 We exclude 25,108 dyads to include lagged dependent variables, 39,336 dyads due to missing geographic data, and
170,860 observations due to missing status data. Restricting the sample to 178 firms effectively limits our analyses
to dyads within the upper three quartiles of the status distribution for AM Law 200 firms. Gross revenue is
positively but imperfectly correlated with status, so our analyses are most representative of the highest status 110 to
160 U.S. law firms in any year.
11
partners hired in a year is one and in 95 percent of cases three or fewer partners are hired within the year.
Table 1 summarizes annual counts of firms, dyads, and lateral hires for our sample.
-----------------------------------------
Insert Table 1 About Here
-----------------------------------------
Logit estimates are biased when the unconditional probability of a binary event is low. So, we
correct coefficient estimates and standard errors to account for the systematic bias attributable to the
dependent variable’s low unconditional mean. Using a rare-events logit specification, we model the
likelihood of firm i hiring a partner from firm j in year t (King and Zeng, 2001). To account for non-
independence of observations attributable to the same dyads appearing multiple times in our sample, we
also cluster errors by dyad. Alternatively, clustering errors by hiring firm, by source firm, or
simultaneously by hiring firm and source firm yields results consistent with those we report.
Independent variables. The key independent variables are status and profitability differentials
between hiring firm i and source firm j. Our status measure is each firm’s prestige score from Vault.com’s
annual law firm rankings. Annually, Vault identifies the 100 most prestigious law firms in the U.S. based
on interviews with lawyers, industry news, prior rankings of law firms, and surveys of thousands of
associates working at these firms. Thousands of associates rate firms, on a scale of 1 to 10, based on how
prestigious they believe working for each firm would be. Respondents do not rate their own firms and are
asked not to rate firms with which they are not familiar. Rankings are based on each firm’s average rating
in these surveys. These scores measure each firm’s generalized labor market status and fit well with
sociological conceptualizations of status as an aggregate of peer attributions (Gould, 2002).
The number of firms included in the top 100 actually varies from year to year and approximately
half of all AM Law 200 firms do not make Vault’s top 100 each year. Rather than restrict our analysis to
only the most prestigious 100 law firms in each year, we obtained prestige scores directly from Vault for
all firms identified in Vault’s survey exercise – including unranked firms. Firm counts range from 126
firms in 2002 to 166 firms in 2010. This data limitation places an upper bound on the number of firms
that can be included in the dyads we analyze; excluded firms may be considered less prestigious than all
12
other firms included in that year. Given the positive empirical correlation between firm prestige and
profitability (0.76), this sampling restriction restricts variance within our sample and renders it less likely
that our analyses identify status or profitability differential effects than if we analyzed all AM Law 200
firms. In robustness checks, we observe similar results when including unrated firms in the analyzed
sample and assigning categorical status ranks over five status tiers instead of continuous status scores.
To estimate the effects of status differentials on the hiring likelihood, we compute status
differential between firm i and firm j as the absolute value of the difference between the Vault prestige
scores for i and j in that year. We also partition this variable into two separate variables based on whether
hiring the hiring firm i or source firm j is higher in status. The variable |Status differentialij|, Si > Sj is the
absolute value of the prestige score difference between hiring firm i and source firm j when the hiring
firm is of higher status. The variable |Status differentialij|, Si < Sj is the absolute value of the status
difference between i and j when the hiring firm is of lower status. Hypothesis 1 predicts negative
coefficients on both of these variables; the coefficient magnitudes indicate whether or not status
differentials exert different influences on the likelihoods of hiring from higher or lower status firms.
Profitability differential is based on profits-per-equity-partner (PPEP), or total firm profits
divided by the number of firm partners with residual claims, a common industry profitability metric that
is obtained from American Lawyer surveys. All else equal, we assume that a partner’s compensation is,
on average, increasing with their firm’s PPEP. We first compute the absolute value of the difference in
PPEP (in U.S. dollars) between hiring firm i and source firm j and then partition this variable into two
separate variables based on whether the hiring firm is more profitable. The variable |Profitability
differentialij|, Pi > Pj is the absolute value of the dollar profitability difference between hiring firm i and
source firm j when the hiring firm is more profitable. The variable |Profitability differentialij|, Pi < Pj is the
absolute value of the dollar profitability difference between hiring firm i and source firm j when the hiring
firm is less profitable. Hypothesis 2 predicts negative coefficients on both of these variables; the
coefficient magnitudes indicate whether or not profitability differentials exert different influences on the
likelihoods of hiring from higher or lower status firms.
13
To test Hypothesis 3 and Hypothesis 4, we create a set of variables indicating the relative status
and profitability of two firms at risk of hiring from one another. One variable identifies dyads in which
the hiring firm i is lower in status but more profitable than the source firm j (Si<Sj and Pi>Pj). A second
variable identifies dyads in which the hiring firm i is higher in status but less profitable than the source
firm j (Si>Sj and Pi<Pj). A third variable identifies dyads in which the hiring firm i is lower in status and
less profitable than the source firm j (Si<Sj and Pi<Pj). The omitted indicator identifies dyads in which
the hiring firm i is higher status and more profitable than the source firm j. To check that our predictions
are insensitive to differential magnitudes, we also test Hypotheses 3 and 4 using continuous measures.
Control variables. Dyad-level covariates account for factors other than relative status and
profitability that may affect the baseline likelihood of lateral hiring. The number of employment
opportunities a firm offers and the number of times a firm’s employees are hired by competitors are both
functions of employee headcount. Firms of different sizes are also likely to differ in terms of the legal
work they perform and the clients they serve (Heinz, et al., 2001). To account for baseline propensities of
individuals to transition from one firm to another, we compute size differential as the absolute value of
the difference in total number of lawyers (in 100s) employed by hiring firm i and source firm j.
Two variables account for major organizational events that influence both the labor market
supply of a firm’s partners and the firm’s demand for competitors’ partners. Dissolved is an indicator
variable that equals 1 when either hiring firm i or source firm j dissolves in a given year and 0 otherwise.
Lawyers employed by firms that dissolve are more likely to be hired by other firms than lawyers whose
firms survive the entire year and firms about to dissolve probably curtail hiring to survive. Merged is an
indicator variable taking a value of 1 when either the hiring firm i or source firm j merges with another
firm in a given year and 0 otherwise. A merger may indicate either future firm growth plans or,
conversely, the culmination of a firm growth plan; both influence the likelihood of lateral hiring.
Geographic distance influences search costs for potential employers and candidates and
adjustment costs imposed by changing employers. We first compute the spherical distances between each
of firm i’s offices and each of firm j’s offices based on their respective longitudes and latitudes (Sorenson
14
and Audia, 2000). Using data on firm-office headcount obtained from the National Law Journal’s annual
survey of the 250 largest U.S. law firms, we weight each distance by the product of the numbers of
lawyers in the two offices, take the sum of all weighted distances, and divide this by the product of the
total number of lawyers in the two offices. The resulting mean geographic distance variable is a
headcount-weighted measure of the average geographic distance, in hundreds of miles, between a
randomly chosen firm i lawyer and a randomly chosen firm j lawyer.
The variable Si > Sj takes a value of 1 when hiring firm i is higher in status and 0 when the source
firm j is higher in status. This variable accounts for baseline tendencies of partners to move from lower-
status to higher-status firms. Partitioning the status differential variable based on whether firm i is higher
or lower status is equivalent to specifying an interaction term between the status differential and this
indicator variable. Hence, this indicator must be included in all models that partition status differential in
order to obtain accurate coefficient estimates. Similarly, Pi > Pj is an indicator variable that takes a value
of 1 when hiring firm i is more profitable and 0 when source firm j is more profitable. This variable
accounts for baseline tendencies of partners to move between more or less profitable firms.
Two variables reflect the relative position of the two firms in the status and profitability
distributions, respectively. Average status is the average of hiring firm i’s status and source firm j’s status.
This variable captures any potential differences in the baseline likelihood of a lateral transition when the
firms are high in status compared to when both firms are low in status3. Similarly, average profit is the
average of hiring firm i’s PPEP and source firm j’s PPEP.
Because growing and contracting organizations differ in their propensities to hire or lose
employees (Freeman and Hannan, 1975), we include four control variables to account for the propensities
of firm i and firm j to hire or lose partners each year. We include the previous year’s counts of partner
hires by i, departures from i, hires by j, and departures from j. We include two more specific control
variables to account for prior movement of partners between i and j, because such hires may enable
3 Note that because we estimate models with upward and downward status differentials in our analyses, it is not
possible to also include variables for firm i status and firm j status.
15
employers to learn about potential candidates or candidates to learn about potential employers. Firms
might also hire individuals from the same competitors from which it hired in the past, based on co-worker
complementarities (Eisenhardt and Schoonhoven, 1990). Therefore, we include number of partner hires
by firm i from firm j in the prior year and departures from firm i to firm j in the prior year. These
variables also serve as controls for otherwise unobserved differences across dyads.
Areas of legal practice vary across firms. For most firms in our sample, we collected data on firm
practice areas from the annually published Vault Guide to the Top Law Firms. We then constructed a
measure of practice area overlap between i and j as follows:
∑
∑ ∑ ∑
where takes a value of 1 if firm i is in practice area k and 0 otherwise, and likewise takes a value
of 1 if firm j is in practice area k and 0 otherwise. The numerator counts the number of practice areas
common to both firms, and the denominator normalizes by the number of practice areas that are present in
either firm. In other words, this variable is a Jaccard coefficient that indexes the intersection of two sets as
a proportion of their union, taking a value of 1 when two sets are identical.4 In our sample, practice
overlap is roughly normally distributed (mean = 0.29; s.d. = 0.13) and exhibits near-zero correlations with
the dyadic status (-0.07) and profitability (-0.09) differentials. These figures suggest that status and
profitability differences are unlikely to absorb the effects of unmeasured differences in practice areas.
But, for all dyads for which we obtained practice area for both firm i and firm j (87 percent of all dyads)
we subject our key findings to accounting for practice area overlap.
Dyadic observations are not independent because firms appear in multiple dyads. We include
Lincoln’s (1984) autocorrelation control variable, which for a given dyad of hiring firm i and source firm
j is the mean value of the dependent variable across all other dyads in which either i or j appears. This
4 This parsimonious control variable is preferable to manually coding practice areas for tens of thousands of lawyers
for each year in order to construct a more fine-grained measure of headcount-weighted, practice area overlap.
16
variable controls for firm-level tendencies to hire or lose lawyers each year.5 All hiring models include
unreported year fixed effects to account for secular trends in lateral hiring; no constant is reported because
the constant represents the omitted year’s effect. Table 2 presents descriptive statistics and correlations
for all variables included in the analyses.
-----------------------------------------
Insert Table 2 About Here
-----------------------------------------
Figure 2 depicts conditional probabilities of hiring from a more profitable competitor and losing
an employee to a lower status competitor – comparisons relevant to Hypotheses 3 and 4. These
probabilities are obtained from summary statistics and not from multivariate regression coefficients. A
few observations are noteworthy. First, consistent with a positive correlation between organizational
status and profitability (in our sample, the firm-level pairwise correlation is 0.76), in the vast majority of
at-risk dyads the hiring firm i is, relative to source firm j, either (1) higher in status and higher in
profitability or (2) lower in status and lower in profitability. The light grey rectangles represent such
dyads; note the much larger dyad count (64, 289 vs. 17,383 or 17,060). Although fewer dyads deviate
from the positive status-profitability correlation (the black rectangles), the rate of lateral hiring is
significantly more likely when one differential is positive and one is negative than when both profitability
and status differentials are positive or both are negative.
Second, consistent with a status-based hiring advantage, the probability of hiring an employee
from a more profitable competitor is significantly greater if the hiring firm is higher in status than if the
hiring firm is lower in status than the competitor (2.3% vs. 2.1%). Third, consistent with a status-based
retention advantage, the probability of losing an employee to a lower status competitor is significantly
greater if the competitor is more profitable than if the competitor is less profitable than the source firm
(3.2% vs. 2.1%). These comparisons suggest that high status organizations have labor market advantages
in hiring and retaining employees; our dyadic analyses account for alternative explanations.
-----------------------------------------
5 The reported results are robust to computing this variable for the entire period of observation instead of annually.
17
Insert Figure 2 About Here
-----------------------------------------
Results
Table 4 reports results of rare-event logit models of the likelihood that firm i hires a partner from
firm j in year t. Model 1 is the baseline model specification with only control variables. Several variables
increase the baseline likelihood of lateral hiring. The likelihood increases with similarity in size and with
geographic proximity. The likelihood that a firm i hires from firm j increases with the number of partners
in the prior year who departed firm i, departed firm j, moved from firm j to firm i, and moved from firm i
to firm j. The higher the mean status of the dyad, the more likely it is that one firm hires from the other.
Finally, the higher the dyad’s mean profitability the less likely it is that one firm hires from the other.
-----------------------------------------
Insert Table 4 About Here
-----------------------------------------
Model 2 includes three variables to test for effects of status differentials on the likelihood of firm
i hiring from firm j: (1) the absolute status differential when the hiring firm is higher in status, (2) the
absolute status differential when the hiring firm is lower in status, and (3) a variable that indicates
whether the hiring firm is higher in status. Consistent with Hypothesis 1, the larger the status difference
between two firms, the lower the likelihood that one will hires the other. This effect is true for both
negative and positive status differentials.
Model 3 includes three variables to examine effects of profitability differentials on the likelihood
of firm i hiring from another firm j: (1) the absolute profitability differential when the hiring firm is more
profitable, (2) the absolute profitability differential when the hiring firm is less profitable, and (3) a
variable that indicates whether the hiring firm is more profitable. Consistent with Hypothesis 2, the larger
the profitability difference between two firms, the less likely it is that one firm hires from the other. This
effect is true for both negative and positive profitability differentials.
We then examine hiring and retention advantages of status. Model 4 provides a test of
Hypothesis 3. In Model 4, hiring from more profitable competitors corresponds to the variables indicating
that Pi < Pj. Hypothesis 3 predicts that higher status firms are more likely to hire from more profitable
18
competitors than are lower status firms. This implies that the variable for [Si>Sj , Pi<Pj] should have a
larger coefficient than the variable for [Si<Sj , Pi<Pj]. The coefficient and error estimates for these two
variables support Hypothesis 3. The variable for [Si>Sj , Pi<Pj] has a positive coefficient that is
significantly different from the baseline, while [Si<Sj , Pi<Pj] is not significantly different from the
baseline. The difference between the two coefficients is statistically significant (p<0.001), implying that a
firm is more likely to hire from a more profitable competitor when the firm’s status exceeds the
competitor’s status than when the competitor’s status exceed the hiring firm’s status.
Hypothesis 4 predicts that the probability of losing employees to a lower status competitor is
more likely when the competitor is more profitable. This implies that the variable for [Si<Sj , Pi>Pj]
should have a larger coefficient than the variable for [Si<Sj , Pi<Pj]. Support for this prediction is found in
the coefficient and error estimates for these two variables in Model 4. The variable for [Si<Sj , Pi>Pj] has
a positive coefficient that is significantly different from the baseline, while [Si<Sj , Pi<Pj] is not
significantly different from the baseline. The difference between the two coefficients is statistically
significant (p<0.001). This implies that a firm is more likely to lose employees to a lower status
competitor when the competitor is more profitable than when the competitor is less profitable.
In Model 5, we further probe Hypotheses 3 and 4. As prior models indicate, firms typically do not
hire from more profitable competitors or lose employees to lower status competitors. In other words,
profitability and status differentials have negative marginal effects on the likelihood of hiring. But,
Hypothesis 3 suggests that status helps hiring firms overcome negative profitability differentials.
Hypothesis 4 suggests that although profitability helps firms overcome negative status differentials, lower
status competitors must be – atypically – more profitable than a higher status competitor to hire an
employee away from that competitor. To test these implications, we examine whether or not the marginal
effects of status and profitability differentials differ across the conditions represented in Figure 2. If
higher status allows a firm to hire from more profitable competitors than it could otherwise, then the
effect of |Profitability differentialij|, Pi < Pj , which is typically negative, should be weaker when Si > Sj
than when Si < Sj. The estimates in Model 5 suggest that this is the case. The variable |Profitability
19
differentialij|, Pi < Pj , Si > Sj has no significant effect, while the variable |Profitability differentialij|, Pi <
Pj , Si < Sj has a significant negative effect.
Using coefficients from Model 5 in Table 4, Figure 3 plots the marginal effect of source firm j’s
higher profitability on the likelihood that firm i hires from firm j. Dark black lines represent all dyads in
which hiring firm i is lower in status than source firm j; light gray lines represent dyads in which i is
higher in status than j. Solid lines represent the estimated marginal effect, bounded by a dashed line 95
percent confidence interval. When firm i is lower in status than firm j, the confidence interval for the
marginal effect of j’s profitability differential is below zero, implying a significant negative effect on the
likelihood of hiring. But, when firm i is higher in status than j, the confidence interval is not statistically
different from zero except when j is much, much more profitable than i. At almost all points the
confidence intervals do not intersect, indicating that the marginal effect of j’s profitability differential is
significantly stronger when firm i is lower in status than when firm i is higher in status.
To put these effects in context, a firm’s probability of hiring from a higher status competitor
decreases by 10 percent for every $100,000 more in profits per equity partner that the competitor makes
relative to the firm6. Conversely, for higher status firms, the probability of hiring from a more profitable
competitor decreases by only 0.05 percent for every $100,000 more in profits per equity partner that the
competitor makes relative to the firm. The confidence intervals imply that when higher status firms can
just as easily hire from more profitable competitors that make up to $1 million more in profits per equity
partner as they can from equally profitable competitors. Consistent with a status-based hiring advantage,
high status firms are more likely to hire more profitable competitors’ employees than are low status firms.
-----------------------------------------
Insert Figure 3 About Here
-----------------------------------------
Model 5 also examines retention advantages of status. If lower status competitors must be more
profitable in order to hire a higher status firm’s employees, then the effect of |Status differentialij|, Si < Sj ,
6 The marginal effect of a competitor’s profitability differential on probability of hiring is -0.002 per $100,000 in
profits per equity partner. A change of -0.002 represents a 10 percent decrease relative to the unconditional
probability of hiring, which is 0.2.
20
which is typically negative, should be weaker when Pi > Pj than when Pi < Pj. The estimates in Model 5
suggest that this is so. The variable |Status differentialij|, Si < Sj and Pi > Pj has no significant effect, while
the variable |Status differentialij|, Si < Sj and Pi < Pj has a significant negative effect.
Using coefficients from Model 5 in Table 4, Figure 4 plots the marginal effect of a firm’s greater
status over a competitor on the firm’s likelihood of losing employees to that competitor. Note that in this
case the competitor is the hiring firm, and, therefore, denoted by subscript i. Dark black lines represent
cases in which i is less profitable than j and light gray lines represent cases in which i is more profitable
than j. Solid lines represent the estimated marginal effect, bounded by a dashed line 95 percent
confidence interval. When a competitor is less profitable than a firm, the confidence interval for the
marginal effect of the firm’s status differential over the competitor is below zero, implying a significant
negative effect on likelihood of losing employees to that competitor. But, when a competitor is more
profitable than the firm, the confidence interval includes zero, implying no significant effect. At all
points, the confidence intervals do not intersect, indicating that the marginal effect of a firm’s positive
status differential is significantly stronger when competitors are less profitable.
To put these effects in context, when a competitor is less profitable than a firm, the firm’s
probability of losing an employee to that competitor decreases by approximately 40 percent for every 1
unit higher in status than the competitor (approximately 25 ranks higher).7 Conversely, the probability of
losing an employee to a more profitable competitor is insensitive to the dyad’s status differential.
Consistent with a status-based retention advantage, high status firms typically lose employees to lower
status competitors only if those competitors are, atypically, also more profitable.
-----------------------------------------
Insert Figure 4 About Here
-----------------------------------------
Robustness checks. We check the robustness of our results to alternative analytical decisions.
One concern is that status and profitability differentials might absorb the effects of unmeasured
7 The marginal effect of a firm’s status differential on probability of losing an employee is -0.008 per 1 unit in status.
A change of -0.008 represents a 40 percent decrease relative to the unconditional probability of hiring, which is 0.2.
21
differences in practice areas between firms, which could affect the likelihood of hiring. In Model 6 we
estimate the fully specified model on the sub-sample of dyads for which we have practice area
information for both i and j (87 percent of all dyads). Greater practice area overlap does indeed have a
significant positive effect on the likelihood of hiring. But, the coefficient estimates for status and
profitability differentials are virtually identical in specifications that do and do not include this overlap
variable. Given the various lagged dependent variables already included in the specification, this seems
unsurprising. But, our findings seem insensitive to practice differences across dyads.
Another possible concern is non-independence of observations, which potentially affects our
sample in two ways. First, dyadic differentials may be correlated over time and, therefore, dyads are not
independent. Second, the dyadic nature of the data means that each firm i appears across multiple
observations, and i’s attributes do not vary within year. The same is true for each firm j. If the main
effects of interest are primarily being driven by variation in firm-level attributes – e.g. by just the hiring
firm’s status rather than dyadic differentials – then standard errors could be underestimated because firm-
level attributes are not independent across dyads.
Table 4’s results are conditional on our use of Lincoln’s (1984) method and clustering errors at
the dyad level. In supplementary analyses, we instead clustered on firm i and firm j and found results
consistent with those reported here. More conservatively, we also used two-way clustering to
simultaneously account for non-independence across all dyads containing firm i and all dyads containing
firm j (Petersen, 2009). When observations are not independent, two-way clustering simultaneously on
multiple non-nested groupings produces the largest standard error estimates in dyadic data and therefore
the most conservative coefficient estimates. Our main results are virtually identical across all clustering
approaches.8 Based on these analyses and the descriptive statistics, we are confident that our results are
not merely attributable to the non-independence of observations.
Another possible concern is that we have treated status and profitability differentials between
firms as varying continuously when categorical differences might be more relevant. Although our
8 These results are unreported to conserve space but are available from the authors.
22
hypotheses are supported with both dichotomous and continuous differential measures in Table 4, we
probed this issue more extensively. Specifically, we divided firms into five status tiers and five
profitability tiers, where the highest tier consists of the top-25 firms, the next tiers contain firms ranked 26
to 50, 51 to 75, 76 to 100, and the last tier contains firms ranked below than 100 (including unranked
firms). This approach accounts for the possibility that the Vault prestige scores are overly precise or that
the AM Law profitability figures are extensively managed by firms on a year-to-year basis.
In Model 7 of Table 5 we replace the continuous status differentials with dummy variables,
indicating whether firm i is in a higher status tier (Si > Sj) or lower status tier than firm j (Si < Sj). The
excluded category represents dyads in which the two firms are in the same status tier (Si = Sj). We
similarly replace the continuous profitability differentials with dummy variables, indicating whether firm
i is in a higher profitability tier (Pi > Pj) or a lower profitability tier than firm j (Pi < Pj). The excluded
category includes dyads in which the two firms are within the same profitability profit tier (Pi = Pj). The
coefficients for all four dummy variables are negative and significantly different from the baseline,
supporting Hypotheses 1 and 2. Firms are less likely to hire from competitors in higher or lower
profitability tiers than from competitors in the same profitability tier. Firms are also less likely to lose
employees to competitors in lower or higher status tiers than to competitors in the same status tier.
-----------------------------------------
Insert Table 5 About Here
-----------------------------------------
Model 8 examines the implications of tiers for hiring and retention advantages. Here, the
excluded category consists of dyads in which the two firms are in the same profitability and same status
tier (Si=Sj , Pi=Pj). Hypothesis 3 implies that the variable for Si>Sj , Pi<Pj should have a larger coefficient
than the variable for Si<Sj , Pi<Pj. Support for this prediction can be easily seen in the coefficient and
error estimates for these two variables in Model 8. The difference is statistically significant (p<0.001). A
firm is more likely to hire from a competitor in a higher profitability tier when the hiring firm is in a
higher status tier. Hypothesis 4 implies that the variable for Si<Sj , Pi>Pj should have a larger coefficient
23
than the variable for Si<Sj , Pi<Pj. Support for this prediction is seen in the coefficient and error estimates
for these two variables in Model 8. The difference between the two coefficients is statistically significant
(p<0.001). A firm is more likely to lose employees to a competitor in a lower status tier when the
competitor is in a higher profitability tier. These results further support Hypotheses 3 and 4.
Consistent with our results using continuous status and profitability differentials, we also find that
being in a higher status tier weakens the negative marginal effect of the profitability differential on a
firm’s likelihood of hiring from a higher profitability tier. When a firm is lower in status, it is less likely
to hire from a competitor in a higher profitability tier than a competitor in the same profitability tier. But
when a firm is in a higher status tier, it is no less likely to hire from a competitor in a higher profitability
tier than to hire from a competitor in the same profitability tier9. These results further support our
argument that high status firms have a hiring advantage over lower status competitors. We also find that a
competitor has to be in a higher profitability tier to weaken the negative marginal effect of status
differential on its likelihood of hiring from a higher status tier. When a competitor is less profitable, it is
less likely to hire from a firm in a higher status tier than a firm in the same status tier. But when a
competitor is in a higher profitability tier, it is no less likely to hire from a firm in a higher status tier than
to hire from a firm in the same status tier10
. These results further support our argument that high status
firms have a retention advantage over lower status competitors.
Concluding Discussion
9 The coefficient for Si<Sj , Pi<Pj is significantly smaller than the coefficient for Si<Sj , Pi=Pj (p<0.001) but the
coefficient for Si>Sj , Pi<Pj is not significantly different from the coefficient for Si>Sj , Pi=Pj (p=0.66). There is,
therefore, a significant difference between higher and lower status firms in the degree to which the likelihood of
hiring decreases when hiring from a higher versus same profitability tier. The difference given by Si<Sj , Pi<Pj -
Si<Sj , Pi=Pj is significantly larger than the difference given by Si>Sj , Pi<Pj - Si>Sj , Pi=Pj (p<0.001). 10
The coefficient for Si<Sj , Pi<Pj is significantly smaller than the coefficient for Si=Sj , Pi<Pj (p<0.001) but the
coefficient for Si<Sj , Pi>Pj is not significantly different from the coefficient for Si=Sj , Pi>Pj (p=0.53). There is,
therefore, a significant difference between more and less profitable competitors in the degree to which the likelihood
of hiring decreases when hiring from a higher status versus same status tier. The difference given by Si<Sj , Pi<Pj -
Si=Sj , Pi<Pj is significantly larger than the difference given by Si<Sj , Pi>Pj - Si=Sj , Pi>Pj (p<0.001).
24
Departing from the established product market advantages of organizational status, we argued
that status is also advantageous in labor markets because employer status is a basis for non-pecuniary
employment benefits. By integrating the economic theory of equalizing differences with the status based
model of market competition, we contribute to organizational theory the insight that organizational status
aids in hiring and retaining employees. We demonstrate two labor market advantages of organizational
status. First, high status firms are more likely to hire employees from more profitable competitors than
low status firms are. Second, high status firms are most likely to lose employees to lower status
competitors when those competitors are atypically more profitable. Both findings run counter to the
central tendency of organizations to hire individuals from competitors of similar profitability and status.
Several conditions bound the generalizability of our findings. First, the theorized trade-offs
between pecuniary and non-pecuniary benefits are probably most advantageous in human capital
intensive industries (e.g., professional services) in which employees represent the primary production
technology and large capital investments are atypical. Status is, of course, particularly valuable in such
industries because quality is difficult to verify ex ante. Second, the legal profession is characterized by
increasing mobility and a common method for delineating employee contributions to organizational
profitability (i.e., a partner’s “book of business”). Future research might test our arguments in settings
where individuals’ contributions are not so readily delineated.
Third, we explicitly assumed – based on prior research – that pecuniary and non-pecuniary
benefits vary, respectively, with organizational profitability and status. Although we cannot observe
employment terms in our data, recent work is consistent with our arguments (Bidwell, et al., 2013). Our
findings indicate that status is advantageous in hiring from more profitable competitors and that
profitability is advantageous in hiring from higher status competitors. Future studies might estimate the
implied exchange rate between pecuniary benefits and employer by analyzing job offer data for senior
employees who, if hired, seem most likely to enhance employer status and profitability.
More broadly, our findings are relevant for the study of individual status attainment. Our results
suggest that most inter-organizational transitions offer individuals only marginal improvements in
25
pecuniary and non-pecuniary benefits because the baseline tendency is to move between organizations of
similar status and profitability. Moreover, a substantial increase in employer status is likely to necessitate
compromising on pecuniary benefits because high status organizations are more likely to hire individuals
from more profitable competitors than low status organizations are. Conversely, a substantial increase in
pecuniary benefits is likely to necessitate compromising on intraprofessional status because more
profitable organizations are more likely to hire employees from high status competitors than low status
competitors are. Based on these findings, individual status attainment studies might consider more
broadly the trade-offs that individuals accept to work for particular employers (e.g., Stern, 2004).
At the organization and industry levels, our results suggest that lateral hiring stabilizes industry
status hierarchies. Low status organizations that exhibit atypically high profitability are most likely to hire
employees away from higher status but less profitable competitors. But, these organizations are also at
greatest risk of losing employees to those competitors. When a low status organization is most likely to
capitalize on atypically high profitability, it is also most vulnerable to losing the employees responsible
for producing profits. Conversely, when a high status organization is most likely to lose employees due
to low profitability that organization is also likely to shore up profitability by hiring employees away
from more profitable competitors. These scenarios imply that high status organizations likely restore
profitability to levels consistent with their status via hiring while low status organizations find it difficult
to sustain abnormally high profitability that might enhance their status. Future research might, therefore,
investigate hiring from competitors as a mechanism that reproduces the frequently-observed positive
correlation between organizational status and profitability. Such inquiries would answer recent calls to
investigate how dyadic exchanges shape status hierarchies (Lynn, Podolny, and Tao, 2010; Sauder, Lynn,
and Podolny, 2012).
Our arguments imply that hiring aids organizational status attainment, a phenomenon that has
received much less attention than individual status attainment has (see Lin, 1999 for a review) but seems
to be attracting growing scholarly interest. For example, Tortoriello, et al., (2011) theorize status-seeking
organizational behavior but, empirically, analyze how such objectives condition inter-organizational
26
exchange, not intra-organizational status change. Cowen (2012) ambitiously models changes in status for
banks but restricts her sample to 123 banking merger and acquisitions instead of the entire banking status
hierarchy. Others analyze larger samples but model status change based on the deference an actor receives
from alters (Bothner, Smith, and White, 2010; Askin and Bothner, 2012). Surprisingly, no prior studies
directly address the question of whether or not strategic efforts to attain greater status actually produce
benefits in excess of attainment costs.
Hiring seems one mechanism by which organizations gain and lose status. Alternatively,
preparing employees for positions at higher status employers might also influence status changes through
status transfer (Baty, et al., 1971; Podolny and Phillips, 1996). Consider that the “entrepreneurs as
organizational products” literature demonstrates that some organizations prepare employees for
entrepreneurship better than others (Freeman, 1986; Burton, et al., 2002). Similarly, many individuals
believe that some organizations are better than others at preparing employees for future attainment (e.g.,
Bidwell, et al., 2013). Other work documents how employees’ prior employment affiliations benefit
employers (e.g., Carnahan and Somaya, 2013). In contrast to extant research that treats lateral hiring as
potentially deleterious to organizations (e.g., Coff, 1997; Hillman, 2002; Broschak, 2004; Gardner, 2005),
future work might consider that losing employees to higher status competitors might be status-enhancing
for both individual and organization.
27
References
Abbott, A. 1981. Status strain in the professions. Amer. J. Sociol. 86(4) 819-835.
Abowd, J.M., F. Kramarz, D.N. Margolis. 1999. High wage workers and high wage firms. Econometrica,
67(2): 251-333.
Almeida, P., B. Kogut. 1999. Localization of knowledge and the mobility of engineers in regional
networks. Management Sci. 45(7) 905–917.
Askin, N. S., M. S. Bothner 2012. Peer effects in tournaments for status: Evidence from rank dynamics of
U.S. colleges and universities. Working paper, U. of Chicago.
Baty, G. B., W. M. Evan, T. W. Rothermel 1971. Personnel flows as interorganizational relations. Admin.
Sci. Quart. 16(4) 430-43.
Benjamin, B. A., J. M. Podolny 1999. Status, quality, and social order in the California wine industry,
1981-1991. Admin. Sci. Quart. 44(3) 563-589.
Bidwell, M., F. Briscoe 2010. The dynamics of interorganizational careers. Organ. Sci. 21(5) 1034-1053.
Bidwell, M., S. Won, R. Barbulescu, E. Mollick. 2013. ‘I used to work at Goldman Sachs!’ How firms
benefit from organizational status in the market for human capital. Working paper, U. of
Pennsylvania.
Bielby, W. T., D. D. Bielby. 1999. Organizational mediation of project-based labor markets: Talent
agencies and the careers of screenwriters. Amer. Soc. Rev.¸ 64(1): 64-85.
Bothner, M. S., E. B. Smith, H. C. White 2010. A model of robust positions in social networks. Amer. J.
of Sociol., 116(3): 943-942.
Broschak, J. P. 2004. Managers’ mobility and market interface: The effect of managers’ career mobility
on the dissolution of market ties. Admin.Sci.Quart. 49(4) 608–640.
Burton MD, Sørensen JB, Beckman CM. 2002. Coming from good stock: Career histories and new
venture formation. Res. Sociol. Organ. 19: 229-262.
Carnahan, S., D. Somaya. 2013. Alumni effects and relational advantage: The impact on outsourcing
when a buyer hires employees from a supplier’s competitors. Acad. Management J. 56(6): 1578-1600.
Chung, S., H. Singh, K. Lee 2000. Complementarity, status similarity, and social capital as drivers of
alliance formation. Strat. Management J. 21(1) 1-22.
Coff, R. 1997. Human assets and management dilemmas: Coping with hazards on the road to resource-
based theory. Acad. of Management R. 48(2), 374-402.
Coleman, J.S. 1990. Foundations of Social Theory. Cambridge, MA: Harvard Univ. Press.
Cowen, A.P. 2012. An expanded model of status dynamics: The effects of status transfer and interfirm
coordination. Acad. Management J. 55(5) 1169-1186.
28
Deckop, J. R. 1988. Determinants of chief executive officer compensation. Ind. and Labor Rel. Rev.,
41(2): 215-226.
Dokko, G., L. Rosenkopf. 2010. Social capital for hire? Mobility of technical professionals and firm
influence in wireless standards committees. Organ. Sci. 21(3) 677-695.
Eisenhardt, K. M., C. B. Schoonhoven 1990. Organizational growth: Linking founding team, strategy,
environment, and growth among U.S. semiconductor ventures, 1978-1988. Admin. Sci. Quart. 35(3)
504-529.
Elsbach, K.D., M.A. Glynn. 1996. Believing your own PR: Embedding identification in strategic
reputation. Adv. in Strat. Manag., 13: 65–90.
Fershtman, C., Y. Weiss. 1993. Social status, culture and economic performance. The Econ. J. 103(419)
946-959.
Frank, R. H. 1985. Choosing the Right Pond: Human Behavior and the Quest for Status. New York:
Oxford University Press.
Freeman, J. H. 1986. Entrepreneurs as organizational products: Semiconductor firms and venture capital
firms,” Adv. in the Study of Entrep., Innov., and Econ. Growth, 1 33-52.
Freeman, J. H., M. T. Hannan 1975. Growth and decline processes in organizations. Amer. Sociol. Rev.
40(2) 215–228.
Galanter, M., T. Palay 1991. Tournament of Lawyers. Chicago: University of Chicago Press.
Gardner, T. 2005. Interfirm competition for human resources: Evidence from the software industry.
Acad. Management J. 48(2) 237-256.
Gerhart, B., G. T. Milkovich. 1990. Organizational differences in managerial compensation and financial
performance. Acad. Management J. 33(4) 663-691.
Gould, R. V. 2002. The origins of status hierarchies: A formal theory and empirical test. Amer. J. Sociol.
107(5) 1143-78.
Goode, W. J. 1978. The Celebration of Heroes: Prestige as a Social Control System. Berkeley: Univ. of
California Press.
Groysberg, B., L.E. Lee, A. Nanda 2008. Can they take it with them? The portability of star knowledge
workers’ performance. Management Sci. 54(7) 1213-1230.
Heinz, J. P. 2009. When Law Firms Fail. Suffolk Univ. Law Rev. 43 67-78.
Heinz, J. P., R. L. Nelson, R. L. Sandefur, and E.O. Laumann. 2005. Urban Lawyers: The New Social
Structure of the Bar. Chicago: University of Chicago Press.
Higgins, M. C., R. Gulati 2003. Getting off to a good start: The effects of upper echelon affiliations on
underwriter prestige. Organ. Sci. 14(3) 244–263.
29
Higgins, M. C., R. Gulati 2006. Stacking the deck: The effects of top management backgrounds on
investor decisions. Strat. Management J. 27(1) 1-25.
Hillman, R. W. 2002. The hidden costs of lawyer mobility: Of law firms, law schools, and the education
of lawyers. Kentucky Law J. 91: 299-310.
Hogan, R., J. Hogan 1991. Personality and status. In D. G. Gilbert, J. J. Connolly (eds.), Personality,
Social Skills, and Psychopathology: An Individual Differences Approach: 137-154. New York:
Plenum Press.
Hsu, D. H. 2004 What do entrepreneurs pay for venture capital affiliation? J. of Finance 59(4) 1805-
1844.
Jensen, M. 2006. Should we stay or should we go? Accountability, status anxiety, and client defections.
Admin. Sci. Quart. 51(1) 97-128.
Jensen, M. and Roy, A. 2008. Staging exchange partner choices: when do status and reputation matter?
Academy of Management Journal. 51(3): 495-516.
King, G., L. Zeng 2001. Logistic regression in rare events data. Pol. Analysis 9 137-163.
Koppel, N. 2013. LeBouef shooting for the stars. The American Lawyer. November 30th. URL last
accessed Feb. 19, 2013: http://blogs.wsj.com/law/2007/11/30/leboeuf-shooting-for-the-stars/
Li, V. 2013. Escape routes. The American Lawyer. February 1st. URL last accessed Feb. 19, 2013:
http://www.americanlawyer.com/PubArticleTAL.jsp?id=1202585323786&thepage=1
Lin, N. 1999. Social networks and status attainment. Annual Rev. Sociol. 25 467-87.
Lincoln, J. R. 1984. Analyzing relations in dyads: Problems, models, and an application to
interorganizational research. Sociol. Methods Research 13(1) 45-76.
Lynn, F.B., J. M. Podolny, L. Tao. 2010. A sociological (de)construction of the relationship between
status and quality. Amer. J. Sociol. 115(3) 755-804.
Major, Lindsey & Africa, LLC. 2012. Partner Compensation Survey: 2012.
McEvily, B.. J. Jaffe, M. Tortoriello. 2012. Not all bridging ties are equal: Network imprinting and firm
growth in the Nashville legal industry, 1933–-1978. Organ. Sci., 23(2) 547–563.
Merton, R. K. 1968. The Matthew effect in science. Science, 159(3810) 56-63.
Meyer, J., B. Rowan 1977. Institutionalized organizations: Formal structure as myth and ceremony. Amer.
J. Sociol. 83(2) 340-363.
Petersen, M. A. 2009. Estimating standard errors in finance panel data sets: Comparing approaches.
Rev. Financ. Stud. 22(1): 435-480.
Phillips, D. J. 2001. The promotion paradox: The relationship between organizational mortality and
employee promotion chances in Silicon Valley Law firms, 1946-1996. Amer. J. Sociol. 106(4) 1058-
98.
30
Phillips, D. J., E. W. Zuckerman 2001. Middle-status conformity: Theoretical restatement and empirical
demonstration in two markets. Amer. J. Sociol. 107(2) 379-429.
Podolny, J. M. 1993. A status-based model of market competition. Amer. J. Sociol. 98(4) 829-872.
Podolny, J. M. 1994. Market uncertainty and the social character of economic exchange. Admin. Sci.
Quart. 39(3) 458-83.
Podolny, J. M. 2001. Networks as the pipes and prisms of the market. Amer. J. Sociol. 107(1) 33-60.
Podolny, J. M. 2005. Status Signals. Princeton, NJ: Princeton University Press.
Podolny, J. M., D.J. Phillips 1996. The dynamics of organizational status. Ind. Corp. Change 5(2) 453-
471.
Podolny, J. M., F. Scott-Morton 1999. Social status, entry, and predation: The case of British shipping
cartels, 1879-1929. J. Industrial Economics 47(1) 41-67.
Podolny, J. M., T.E. Stuart, M. T. Hannan 1996. Networks, knowledge, and niches: Competition in the
worldwide semiconductor industry. Amer. J. Sociol. 102(3) 659-89.
Rao, H., R. Drazin 2002. Overcoming resource constraints on product innovation by recruiting talent
from rivals: A study of the mutual fund industry, 1986-94. Acad. Management J. 45(3) 491-507.
Rivera, L. 2012. Hiring as cultural matching: The case of elite professional service Firms. Amer. Sociol.
Rev. 77(6) 999-1022 .
Roberts, P. W., M. Khaire 2009. Getting known by the company you keep: Publicizing the qualifications
and associations of skilled employees. Ind. Corp. Change, 18(1) 77–106.
Roberts, P. W., M. Khaire, C. I. Rider 2011. Isolating the symbolic implications of employee mobility:
Price increases after hiring winemakers from prominent wineries. Amer. Econ. Rev. Papers and
Proceedings, 101(3): 147-151.
Rosen, S. 1986. The theory of equalizing differences. Handbook of Labor Economics. 1. 641-692.
Sauder, M. 2008. Interlopers and field change: The entry of U.S. News into the field of legal education.
Admin. Sci. Quart. 53(2) 209-234.
Sauder, M., F.B. Lynn, J.M. Podolny. 2012. Status: Insights from organizational sociology. Ann. Rev. of
Sociol. 38 267-283.
Sherer, P. D., K. Lee 2002. Institutional change in large law firms: A resource dependency and
institutional perspective. Acad. Management J. 45(1) 102-119.
Somaya, D., I. O. Williamson, and N. Lorinkova. 2008. Gone but not lost: The different performance
impacts of employee mobility between cooperators and competitors. Acad. Management J. 51(5):
936-953.
31
Sørensen, J. B. 1999. Executive migration and interorganizational competition. Soc. Science Res. 28(3)
289-315.
Sørensen, J. B., O. Sorenson 2007. Corporate demography and income inequality. Amer. Sociol. Rev.
72(5) 766-783.
Sorenson, O., P.G. Audia 2000. The social structure of entrepreneurial activity: Geographic
concentration of footwear production in the United States, 1940-1989. Amer. J. Sociol. 106(2) 424-
462.
Stern, S. 2004. Do scientists pay to be scientists? Management Sci. 50(6): 835–853.
Tortoriello, M., V. Perrone, B. McEvily 2011. Cooperation among competitors as status-seeking
behavior: Network ties and status differentiation. Euro. Management J. 29(5) 335-346.
Triedman, J. 2006. To lure top laterals, LeBoeuf Lamb turns rainmaker magnet. The American Lawyer.
February 1st. URL last accessed Feb. 19, 2013:
http://www.law.com/jsp/article.jsp?id=1138701911788
Uzzi, B., R. Lancaster. 2004. Embeddedness and price formation in the corporate law market. Amer.
Sociol. Rev. 69 (3). 319-344.
Wezel, F. C., G. Cattani, J. M. Pennings 2006. Competitive implications of interfirm mobility. Organ.
Sci. 17(6) 691 -709.
Figure 1. Lateral partner hiring by American Lawyer 200 Law Firms, 2000-2009.
Note: The solid blue line depicts the number of lateral partner transitions between one AM Law 200 firm to
another in each year. The dashed red line depicts that number as a percentage of all partner hires involving an
AM Law 200 firm, including hiring from employers not included in the AM Law 200.
0%
10%
20%
30%
40%
50%
60%
70%
0
250
500
750
1,000
1,250
1,500
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Number of lateral hires Percent of all hires
32
Figure 2. Conditional probabilities that i hires from j (Hypotheses 3 and 4).
Note: Both differences depicted above (i.e., 2.1% vs. 2.3% and 2.1% vs. 3.2%) are statistically significant (p <
0.001). Figures are summary statistics; neither is subject to multivariate regression.
Figure 3. Marginal effect of j’s positive profitability differential on likelihood that i hires from j.
Note: Solid lines depict the estimated marginal effect on the predicted likelihood. Dashed lines represent the
upper and lower bounds of the 95 percent confidence interval around the marginal effect. Black lines represent
the marginal effect on predicted likelihoods of hiring when hiring firm i is lower status than source firm j.
Gray lines represent the marginal effect on predicted likelihoods when i is higher status than j.
-0.0025
-0.002
-0.0015
-0.001
-0.0005
0
0.0005
0 2 4 6 8 10 12 14
d p
r(i
hir
es f
rom
j)
/ d
(p
rofi
t(j)
- p
rofi
t(i)
)
Profitability(j)-profitability(i)
$100k
status(i) > status(j) status(i) < status(j)
2.1%2.3%
Hiring firm is lower status(n=64,289).
Hiring firm is higher status(n=17,383).
Conditional probabilities of hiring from a more profitable competitor (H3).
2.1%
3.2%
Hiring firm is less profitable(n=64,289).
Hiring firm is moreprofitable (n=17,060).
Conditional probabilities of losing an employee to a lower status competitor (H4).
33
Figure 4. Marginal effect of j’s positive status differential on likelihood that i hires from j.
Note: Solid lines depict the estimated marginal effect on the predicted likelihood. Dashed lines represent the
upper and lower bounds of the 95 percent confidence interval around the marginal effect. Black lines represent
the marginal effect on predicted likelihoods of hiring when hiring firm i is less profitable than source firm j.
Gray lines represent the marginal effect on predicted likelihoods when i is higher more profitable than j.
Table 1. Sample firms, dyads, and hires by year.
Year Firms Dyads Hires
2002 114 12,882 331
2003 139 19,182 419
2004 144 20,592 619
2005 143 20,306 763
2006 149 22,052 688
2007 155 23,870 994
2008 150 22,350 790
2009 147 21,462 666
All 178 162,696 5,270
-0.012
-0.01
-0.008
-0.006
-0.004
-0.002
0
0.002
0 0.2 0.4 0.6 0.8 1 1.2 1.4
d p
r(i
hir
es f
rom
j)
/ d
(st
atu
s(j)
- s
tatu
s(i)
)
Status(j) - status(i)
profit(i) > profit(j) profit(i) < profit(j)
34
Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1 Organization i hires from competitor j , (0/1) 0.02 0.15 --
2 Size differential (100s of lawyers) 0.0 5.7 -0.01 --
3 i or j dissolved (t-1), (0/1) 0.0 0.04 0.02 0.00 --
4 i or j merged (t-1), (0/1) 0.03 0.16 0.01 0.00 -0.01 --
5 Autocorrelation control (ij ) 0.02 0.01 0.19 -0.03 0.06 0.04 --
6 # of departures from i (t-1) 6.9 7.5 0.06 0.30 0.06 0.05 0.28 --
7 # of hires by i (t-1) 11.4 14.9 0.08 0.31 0.00 0.02 0.37 0.41 --
8 # of departures from j (t-1) 6.9 7.5 0.09 -0.30 0.06 0.05 0.38 0.03 0.01 --
9 # of hires by j (t-1) 11.4 14.9 0.05 -0.31 0.00 0.02 0.24 0.01 0.00 0.41 --
10 # of partners i hired from j (t-1) 0.03 0.30 0.09 0.00 0.01 0.01 0.10 0.05 0.09 0.11 0.04 --
11 # of partners j hired from i (t-1) 0.03 0.30 0.03 0.00 0.01 0.01 0.07 0.11 0.04 0.05 0.09 0.02 --
12 Mean geographic distance (100s of miles) 8.9 4.4 -0.01 0.00 0.02 0.01 0.06 0.03 0.03 0.03 0.03 -0.01 -0.01 --
13 Mean status (ij ) 5.2 1.0 0.05 0.00 -0.02 -0.03 0.16 0.09 0.07 0.09 0.07 0.03 0.03 -0.06 --
14 Mean profit (ij) 9.7 4.3 0.00 0.00 0.00 -0.04 0.00 -0.04 -0.07 -0.04 -0.07 0.00 0.00 -0.12 0.59 --15 S i > S j (0/1) 0.50 0.50 -0.02 0.41 0.00 0.00 -0.05 0.10 0.08 -0.10 -0.08 0.00 0.00 0.00 0.00 0.00 --
16 P i > P j (0/1) 0.50 0.50 0.00 0.20 0.00 0.00 -0.03 -0.02 -0.02 0.02 0.01 0.00 0.00 0.00 0.00 0.00 0.58 --
17 Profit differential (i - j) , ($100,000s) 5.7 5.8 -0.05 0.00 0.00 -0.03 -0.11 -0.08 -0.10 -0.08 -0.10 -0.03 -0.03 -0.07 0.33 0.71 0.00 0.00 --
18 Status differential (i - j) 1.6 1.2 -0.06 0.00 -0.01 -0.03 -0.09 -0.06 -0.07 -0.06 -0.07 -0.04 -0.04 -0.06 0.24 0.31 0.00 0.00 0.53 --
19 |Status differential ij |, Si > Sj 0.79 1.2 -0.05 0.42 -0.01 -0.01 -0.11 0.05 0.03 -0.11 -0.10 -0.02 -0.02 -0.03 0.12 0.16 0.69 0.54 0.27 0.51 --
20 |Status differential ij |, Si < Sj 0.79 1.2 -0.02 -0.42 -0.01 -0.01 0.02 -0.11 -0.10 0.05 0.03 -0.02 -0.02 -0.03 0.12 0.16 -0.69 -0.54 0.27 0.51 -0.47 --
21 |Profit differential ij |, Pi > Pj 2.8 5.0 -0.04 0.12 0.00 -0.02 -0.10 -0.10 -0.12 0.01 0.01 -0.02 -0.02 -0.04 0.19 0.42 0.44 0.57 0.58 0.31 0.66 -0.35 --
22 |Profit differential ij |, Pi < Pj 2.8 5.0 -0.02 -0.12 0.00 -0.02 -0.03 0.01 0.01 -0.10 -0.12 -0.02 -0.02 -0.04 0.19 0.42 -0.44 -0.56 0.58 0.31 -0.35 0.66 -0.32
Table 2. Correlations for variables in dyadic hiring analyses (n = 162,696 dyads).
35
Size differential (100s of lawyers) -0.006 † -0.002 -0.004 -0.003 -0.003 -0.003
(0.003) (0.004) (0.003) (0.003) (0.004) (0.004)
i or j dissolved (t-1), (0/1) -0.230 -0.175 -0.124 -0.222 -0.121 -0.091
(0.236) (0.235) (0.238) (0.238) (0.236) (0.234)
i or j merged (t-1), (0/1) 0.102 0.070 0.095 0.105 0.072 0.088
(0.099) (0.100) (0.099) (0.099) (0.099) (0.097)
Autocorrelation control (ij ) 48.9 *** 47.1 *** 47.2 *** 48.6 *** 46.6 *** 46.0 ***
(1.10) (1.14) (1.13) (1.11) (1.15) (1.18)
# of departures from i (t-1) 0.008 *** 0.007 ** 0.007 ** 0.008 *** 0.006 ** 0.006 **
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
# of hires by i (t-1) 0.000 0.000 0.000 0.000 -0.001 -0.001
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
# of departures from j (t-1) 0.014 *** 0.014 *** 0.015 *** 0.015 *** 0.015 *** 0.014 ***
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
# of hires by j (t-1) 0.001 0.000 0.000 0.000 0.000 -0.000
(0.001) (0.001) (0.001) (0.001) (0.001) (0.001)
# of partners i hired from j (t-1) 0.316 *** 0.282 *** 0.295 *** 0.311 *** 0.277 *** 0.226 *
(0.081) (0.083) (0.081) (0.081) (0.083) (0.094)
# of partners j hired from i (t-1) 0.083 *** 0.068 ** 0.072 *** 0.083 *** 0.063 ** 0.069 **
(0.023) (0.021) (0.021) (0.022) (0.021) (0.021)
Geographic distance (100s of miles) -0.038 *** -0.043 *** -0.039 *** -0.038 *** -0.042 *** -0.036 ***
(0.006) (0.005) (0.005) (0.006) (0.005) (0.006)
Mean status (ij ) 0.393 *** 0.385 *** 0.302 *** 0.386 *** 0.333 *** 0.345 ***
(0.032) (0.030) (0.032) (0.031) (0.031) (0.033)
Mean profit (ij) -0.076 *** -0.062 *** 0.008 -0.073 *** -0.015 -0.016
(0.008) (0.008) (0.009) (0.008) (0.010) (0.010)
S i > S j (0/1) 0.085
(0.059)
|Status differential ij |, Si < Sj -0.342 ***
(0.027)
|Status differential ij |, Si > Sj -0.531 ***
(0.031)
P i > P j (0/1) 0.189 ***
(0.055)
|Profitability differential ij |, Pi < Pj -0.076 ***
(0.007)
|Profitability differential ij |, Pi > Pj -0.132 ***
(0.008)
Si < Sj and Pi < Pj (0/1) -0.078 0.221 ** -0.197 *
(0.050) (0.077) (0.080)
Si < Sj and Pi > Pj (0/1) 0.227 *** -0.097 -0.247 **
(0.066) (0.093) (0.096)
Si > Sj and Pi < Pj (0/1) 0.313 *** -0.023 -0.300 ***
(0.060) (0.097) (0.088)
|Status differential ij |, Si > Sj -0.427 *** -0.409 ***
(0.037) (0.039)
|Status differential ij |, Si < Sj and Pi > Pj -0.140 † -0.125
(0.079) (0.080)
|Status differential ij |, Si < Sj and Pi < Pj -0.305 *** -0.309 ***
(0.035) (0.037)
|Profitability differential ij |, Pi > Pj -0.087 *** -0.084 ***
(0.010) (0.010)
|Profitability differential ij |, Pi < Pj and Si > Sj -0.017 -0.019
(0.017) (0.018)
|Profitability differential ij |, Pi < Pj and Si < Sj -0.044 *** -0.041 ***
(0.009) (0.009)
Practice area overlap 0.613 ***
(0.164)
Dyad observations
Log pseudolikelihood
Pseudo R2
Chi-square (df) 5,225.2 (20) 5,464.3 (23) 5,113.7 (23) 5,152.1 (23) 5,346.8 (29)4946.81 (30)
*** p<0.001, ** p<0.01, * p<0.05, † p<0.10
Table 4. Rare event logit models of the likelihood that i hires from j (n = 162,696 dyads).
(6)
-8,887.6
0.141
162,696 162,696 162,696 162,696 162,696 141,992
(1) (2) (3)
-10,382.4
0.124
All models include unreported year fixed effects. Robust standard errors in parentheses; clustered by dyad
-10,216.5 10,261.4
(5)
-10,176.0
0.1410.138
(4)
-10,362.1
0.1260.134
36
Size differential (100s of lawyers) -0.003 -0.003
(0.003) (0.003)
i or j dissolved (t-1), (0/1) -0.284 -0.276
(0.228) (0.229)
i or j merged (t-1), (0/1) 0.162 † 0.152 †
(0.092) (0.092)
Autocorrelation control (ij ) 53.3 *** 53.3 ***
(1.079) (1.086)
# of departures from i (t-1) 0.009 *** 0.010 ***
(0.002) (0.002)
# of hires by i (t-1) 0.000 -0.000
(0.001) (0.001)
# of departures from j (t-1) 0.015 *** 0.015 ***
(0.002) (0.002)
# of hires by j (t-1) 0.002 † 0.001
(0.001) (0.001)
# of partners i hired from j (t-1) 0.383 *** 0.381 ***
(0.059) (0.059)
# of partners j hired from i (t-1) 0.096 ** 0.093 **
(0.032) (0.032)
Geographic distance (100s of miles) -0.049 *** -0.040 ***
(0.005) (0.005)
Mean status tier (ij ) -0.333 -0.318
(0.025) (0.025)
Mean profitability tier (ij) 0.106 *** 0.104 ***
(0.026) (0.025)
S i > S j (≥ 1 tier ) -0.258 ***
(0.052)
S i < S j (≤ 1 tier ) -0.343 ***
(0.054)
P i > P j (≥ 1 tier) -0.410 ***
(0.049)
P i < P j (≤ 1 tier ) -0.481 ***
(0.052)
S i > S j and P i > P j (by tier ) -0.651 ***
(0.062)
Si > S j and P i = P j (by tier ) -0.219 **
(0.076)
S i < S j and P i = P j (by tier ) -0.139 †
(0.078)
S i = S j and P i > P j (by tier ) -0.278 **
(0.085)
S i = S j and P i < P j (by tier ) -0.325 ***
(0.084)
S i < S j and P i > P j (by tier ) -0.344 ***
(0.098)
S i < S j and P i < P j (by tier ) -0.866 ***
(0.064)
S i > S j and P i < P j (by tier ) -0.260 **
(0.094)
Constant (S i =S j , P i =P j ) -3.94 *** -4.05 ***
(0.101) (0.108)
Observations 265,840 265,840
Log pseudo-likelihood
Pseudo R-sqr 0.145 0.146
Chi-square (df) 7,368 (24) 7,301 (28)
Robust standard errors in parentheses; clustered by dyad.
Table 5. Robustness checks: Status and profitability tiers.
All models include unreported year fixed effects.
(7) (8)
*** p<0.001, ** p<0.01, * p<0.05, + p<0.10
-16,561.0 -16,531.0