Up for Review: Unravelling the Link between Formal Evaluations and Performance‐Based Rewards

Date01 January 2019
Published date01 January 2019
DOIhttp://doi.org/10.1111/irel.12223
AuthorTor Eriksson,Jaime Ortega,Rocio Bonet
Up for Review: Unravelling the Link between
Formal Evaluations and Performance-Based
Rewards
*
ROCIO BONET, TOR ERIKSSON and JAIME ORTEGA
We study whether organizations that reward individual performance should give
autonomy or should control how managers evaluate their subordinates. The
normal way to establish control is to formalize the evaluations, so that managers
cannot choose when and how to evaluate. We argue that organizations face a
trade-off because formalization helps reduce biases but also introduces rigidities.
Using linked employeremployee data, we study the link between formal perfor-
mance appraisals and rm nancial performance.
Introduction
An important question for organizations is how much autonomy they should
give managers regarding personnel matters. In this paper we consider in partic-
ular autonomy in conducting a key managerial taskthe performance evalua-
tion of employees. Performance evaluations can have important consequences
for the organization, particularly when they are used to guide administrative
decisions that link employee performance with organizational rewards and
punishments, such as pay, promotions, or dismissals (Rynes, Gerhart, and
Parks 2005). However, we lack empirical evidence that disentangles the effect
of formal evaluations from that of performance-based rewards.
According to academic research, the answer to the control versus autonomy
question is not obvious because of the trade-offs involved. Control is useful to
*The authorsafliations are, respectively, IE Business School, Madrid, Spain. E-mail: rocio.bonet@ie.edu;
Aarhus University, Aarhus, Denmark. E-mail: tor@econ.au.dk; and Universidad Carlos III de Madrid,
Madrid, Spain. E-mail: jortega@emp.uc3m.es. The authors thank the editors Steven Raphael and Chris Rid-
dell and the referees for their interest and valuable comments. They also thank Kasper Kristensen for provid-
ing expert research assistance, and participants at the 3rd Wharton People and Organizations Conference, 1st
Madrid Work and Organizations Workshop, Strategic Management Society 31st Annual International Con-
ference, 73rd Annual Meeting of the Academy of Management, 28th EGOS Colloquium, and the University
of Aberdeen and SIRE Conference on New Research in Performance-Related Pay for their suggestions.
Jaime Ortega acknowledges nancial support from the Spanish Ministry of Economics, Industry and Com-
petitiveness (grants ECO2012-33308 and ECO2015-69615-R). Rocio Bonet acknowledges nancial support
from the European Commission grant WSCA-Grant Agreement No. 239217-7FP-People-IRG.
INDUSTRIAL RELATIONS, DOI: 10.1111/irel.12223. Vol. 58, No. 1 (January 2019). ©2018 The Regents of
the Univers ity of Califo rnia Published by Wiley Periodicals, Inc., 350 Main Street, Malden,
MA 02148, USA, and 9600 Garsington Road, Oxford, OX4 2DQ, UK.
108
align managersdecisions with broader organizational goals, forcing them to
internalize these goals (Bolton and Dewatripont 2013), but it arguably reduces
exibility, limiting their ability to act according to the specic knowledge they
acquire (Baker 1992; Prendergast 2002). In the case of performance evalua-
tions, one of the main concerns with autonomy is managersconscious or
unconscious biases (Castilla 2008). On the other hand, knowledge of their
workers is to a large extent specic, and excessive control may hamper man-
agersability to use this knowledge. In the absence of formalization, it could
be easier for managers to tailor rewards to individual circumstances to better
motivate their subordinates.
For rms, the standard way to exercise control over the evaluations is to
formalize them, so that managers are not completely free to choose when and
how to evaluate. Some examples are rules on the frequency with which man-
agers should evaluate, compulsory use of questionnaires with closed formats,
or restrictions on the scores that managers can give. Though widely spread
and often considered a good practice, formalization is also unpopular among
managers and some academics have questioned the value of formal evalua-
tions. According to one of the main critics, reviewing performance is good
(...) but employees need evaluations they can believe, not the fraudulent ones
they receive. They need evaluations that are dictated by need, not a date on
the calendar. They need evaluations that make them strive to improve, not pre-
tend to be perfect(Culbert 2010: 2).
In practice, and paralleling these mixed theoretical views, rms use different
approaches. Performance appraisal processes seem to be a standard practice,
but there are notable counterexamples. Renowned rms such as Adobe, Juni-
per Networks, Accenture, PwC, SAP, and perhaps even more noteworthy due
to their strong tradition on emphasis on numbers, Goldman Sachs and General
Electric, have been featured in the news for their radical approach to change
their performance appraisal system: getting rid of formal evaluations.
In a recent piece, Cappelli and Conyon (2018) provide descriptive evidence
inconsistent with the criticsidea that performance ratings are uninformative.
This idea is generally based on the argument that managers tend to give very
similar evaluations and do not give poor ratings. Using data from a large retail
U.S. corporation, they nd that there is large variation in the ratings and that
the distribution of the ratings is close to a normal distribution. They also
explore other claims, such as supervisors giving higher ratings to employees
they know for a long time, and nd no evidence consistent with those claims.
Furthermore, Cappelli and Conyon (2018) nd that pay levels are related to
changes in the ratings, i.e., employees who improve are rewarded and those
who regress are penalized. This is consistent with a relational model in which
managers are trying to motivate employees to improve future performance, as
Formal Evaluations and Firm Performance / 109

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT