Performance‐related pay, mental and physiological health

• Published date: 01 January 2024
• Author: Nicole Andelic,Julia Allan,Keith A. Bender,Daniel Powell,Ioannis Theodossiou
• Date: 01 January 2024
• DOI: http://doi.org/10.1111/irel.12334

Industrial Relations. 2024;63:3–25.

|

3wileyonlinelibrary.com/jour nal/irel

Basing employees' s alaries on t heir perfor mance (p erformanc e- related pay or PRP) has tradi-

tionally be en perceived as a payment sys tem with a range of positive outcome s. There is a well-

establishe d link between PRP and h igher productivity th rough incentiviz ation and self- sorting

effects, and c onsequently higher e arnings for the employee and h igher productivity for employ-

ers. Traditionally th en, it has been widely ac cepted that PRP is of benef it to all parties i nvolved

and leads to an a lignment b etween the goal s of the employers and employees. However, in hi s

book The Wealth of Nations, Scotti sh economist Adam Smith ma de the following observation:

“Workmen… when they are libe rally paid by the pi ece, are ver y apt to overwork themselve s

and to ruin t heir health a nd constitution in a few ye ars.” (Smith1937, p. 83). This early obser-

vation suggests t hat PRP m ay have un intended negative consequ ences on workers' health.

Despite th is early observation, it is only i n the past two decades th at researchers have begu n

seriously invest igating this claim empir ically and subsequently f inding a relationship betwe en

Received : 14 September 2022

|

Accepte d: 5 April 2023

DOI: 10.1111/ir el.12334

ORIGINA L ARTICL E

Performance- related pay, mental and physiological

health

NicoleAndelic1,2 | JuliaAllan3 | Keith A.Bender1,2 |

DanielPowell3 | IoannisTheodossiou1

1Economi cs Departm ent, Centre for

Labour Marke t Research, Bu siness Scho ol,

Universit y of Aberdeen, Abe rdeen, UK

2Scottish E xperime ntal Economi cs

Laborator y, Economics De partment,

Busines s School, Univer sity of Aberdee n,

Aberdeen , UK

3Health Psych ology, Institute o f Applied

Health Sci ences, Unive rsity of Aberde en,

Aberdeen , UK

Correspondence

Nicole And elic, Econom ics Depar tment,

Centre for Lab our Market Resea rch,

Busines s School, Univer sity of Aberdee n,

Aberdeen , UK.

Email: nicole.andelic@abdn.ac.uk

Funding information

Economi c and Social Res earch Counc il,

Grant/Award Numb er: ES/R01163X/1

Abstract

Much of the literatu re on performa nce- related pay (PRP)

and poor health rel ies on self- reported dat a, and the re-

lationship is di fficu lt to examine du e to confounding

variables. We exam ine the relationsh ip between PRP

and three g roups of health meas ures using d ata from

the UKHLS: blood pres sure, inf lam mation markers in

blood, and self- r eported he alth. Regression s correct ing

for self- s election bia s and socio- de mographic covar iates

find that PRP contr acts are associat ed with poorer mental

health, high er systolic blood pre ssure, and h igher levels

of fibrinoge n. These fi ndings sugg est that fi rms that us e

PRP may need to implem ent policies t o mitigate agai nst

PRP- related stress.

JEL CLAS SIFICATION

J33, I0, I14

This is an op en acces s article und er the terms of t he Creative Commons Attribution Lic ense, which p ermits us e, distributi on and

reproduct ion in any mediu m, provided th e original work i s properly cite d.

© 2023 The Authors . Industria l Relations publi shed by Wiley Per iodicals LLC o n behalf of Regent s of the Universit y of Californ ia

(RUC ).

4

|

ANDELIC ET AL.

PRP and various mea sures of poor health (Art z & Heywood2015; Bender & The odossiou2014).

These stud ies have to date, al most entirely rel ied on subjective eva luations or recol lections of

injuries , medical c onditions or health. I n contrast, there i s little res earch on PRP and healt h

using biomarkers a s physiological indicator s of health in large- scale survey data. Fur thermore,

there is a well- known sor ting effect whe n workers select into PRP. It is possible then that t he

link bet ween PRP and poor hea lth is due to confoundi ng variables an d sorting, rathe r than

PRP leading to worse he alth.

We address these l imitation s in the cur rent study by using an i ndividual leve l dataset that

includes both l abor market informat ion as well as a nurs e health asse ssment with obje ctive

measures of h ealth. By includi ng contextual variables that a re related to PRP but not dire ctly

related to healt h as instru ments, we are able to st atistical ly correct for the s elf- selection bias

found among PRP workers. In addition to th is important statist ical correction, the avai lability

of a nurse asse ssment of health allows us to exam ine the relationship betwee n PRP a nd health

using more robust and obje ctive measures of health tha n t hose found in the current literat ure.

LITERATURE REVIEW

Although there a re several stud ies that have found a lin k between PRP and he alth in sur vey

data, a wide ra nge of ob served hea lt h outcomes are in cluded in the l iterature suc h as injurie s,

insuranc e premium s, medicat ion use, and sel f- re ported hea lth. Conseque ntly, there is likely

more than one causa l mechanism through wh ich PRP may impact on health and the l iterature

can broadly be de fined into three suc h pathways.

The firs t pathway is based on the a ssociation b etween PRP and injur ies. It is possible t hat

making ea rnings de pendent on per formance in centiviz es individu als to incre ase their r ate

of output, thereby also en couraging r isk- taking behaviors whic h in turn le ad to higher rat es

of injuries a nd accidents at work. Th is is typic ally support ed by industry- specific exampl es.

For example, Freeman and K leiner (2005) f ind that an Ame rican shoe - manufacturing com-

pany that switche d from PRP to fixed s alary contr acts saw a decre ase in workers' comp en-

sation insura nce prem iums, sugge sting that acc idents and injur ies less ened after th e switch.

Comparisons b etween group s demonstrate si milar re sults. For example, a study exa mini ng

long- haul truckers ( Mona co & Williams2000) finds that ac cidents were more frequent among

groups of trucker s who are paid by percentage of revenue i n comparison to those who are pa id

by the mile (18% vs. 13%) and in line with this, S aha et al.(2004) show that workers in fertilizer

production who are pai d by PRP are 1.7– 4.3 time s more likely to have acc idents than the ir

counterpa rts on time- ba sed pay contracts over a 5- year p eriod. Finally, beyond these i ndustry-

specif ic findings, seve ral studies have used nationa lly representative survey d ata and find that

there is a robust li nk between PRP a nd injuries /accidents ac ross industri es and occup ations

(Artz & Heywood2 015; Bender et a l.2 012).

However, PRP is also associate d with health outc omes not direc tly related to inju ries.

For example, Dahl and Pi erce(2020) study Dani sh firm s who switched from f ixed to PRP

contracts and d iscover a 4%– 6% increase in the us e of anti- depressants a nd anxiety m edi-

cation by employees, in dicating that me ntal health was p oorer after the s witch. Bender a nd

Theodossiou(20 14) fi nd that PRP worker s were more likely than others to rep ort issues with

cardiovascu lar health, stom ach/digestive proble ms, and poorer emot ional health. Th is may

be due to a secon d pathway; the conseque nces of working more hour s when employed in a

PRP contract. Prev ious resear ch in the UK has found th at PRP workers are likely to work

1.5 hours more per week t han those on f ixed pay (Bender & T heodossiou 2014 ). Therefore,

there may be a trad e- off of ear nings versus l eisure tim e that could otherwise be sp ent on ac-

tivities that p romote physical and mental hea lth. On the other hand, it may be th at rather than

causing a dec rease in healthy activiti es, the additional hours worked may lead to a n increase