The Longer‐term Impacts of Job Displacement on Labour Market Outcomes in New Zealand

• Author: Dean R Hyslop,Wilbur Townsend
• DOI: http://doi.org/10.1111/1467-8462.12312
• Published date: 01 June 2019
• Date: 01 June 2019

The Longer-term Impacts of Job Displacement on Labour

Market Outcomes in New Zealand

Dean R Hyslop and Wilbur Townsend*

Abstract

This article analyses the longer-term impacts

of job displacements in New Zealand using

linked survey and administrative data. The job

loss rate is low but has substantial and long-

lasting impacts. Displaced workers have 20–25

per cent lower employment rates than non-

displaced workers in the year following

displacement, and still 8–12 per cent lower

after five years. Furthermore, their conditional

earnings and incomes were about 25 per cent

lower in the first year and 15 per cent lower

after five years. The impacts were concentrated

among older workers, larger for those dis-

placed during the great recession, and only

modestly ameliorated by government income

support.

1. Introduction

A dynamic economy with firm entry and exit,

together with a responsive labour market

involving job creation and destruction, is

considered central to economic prosperity

(Davis and Haltiwanger, 1992). Job destruc-

tion is known to typically cause involuntary

unemployment in the short term. However, the

longer-term impacts of job loss are critical to

understanding the adjustment costs borne by

affected workers. Furthermore, there is little

evidence comparing impacts for workers

displaced at different stages of the business

cycle.

1

In this article, we analyse the longer-term

impacts of involuntary job loss over the

business cycle in New Zealand. New Zealand

has relatively low employment protection, no

seniority protection or unemployment insur-

ance, and targeted means-tested income sup-

port from welfare benefits and other transfers.

In addition, New Zealand has comparatively

low unemployment and low rates of job

displacement.

2

Given these features, the

expected labour market impacts of job dis-

placement, and the adequacy of income support

in ameliorating such impacts, are unclear: for

example, displacement may be more selective

in New Zealand and thus displaced workers

may endure worse outcomes, but there may

also be greater job opportunities in a low

unemployment labour market. Our sample

period includes an economic expansion until

the end of 2007, followed by a domestic

recession and the global financial crisis from

2008. This facilitates a comparative analysis of

the impacts of job displacements that occur at

different stages of the business cycle within the

same jurisdiction.

* Hyslop: Motu Economic and Public Policy Research,

Wellington 6142 New Zealand; Townsend: Department of

Economics, Harvard University, Cambridge, MA 02138

United States. Corresponding author: Hyslop, email

<dean.hyslop@motu.org.nz>. This research was sup-

ported by the New Zealand Ministry for Business,

Innovation and Employment (MBIE). We thank the editor,

two referees, Sylvia Dixon, Dave Mar

e, Holly Norton and

seminar participants at MBIE for helpful discussion and

comments. Disclaimer: Access to the anonymised data

used in this study was provided by Statistics New Zealand

in accordance with security and confidentiality provisions

of the Statistics Act 1975, and secrecy provisions of the

Tax Administration Act 1994. The results in this article are

the work of the authors, not Statistics NZ, and have been

confidentialised to protect individuals and businesses from

identification. See Hyslop and Townsend (2017b) for the

full disclaimer.

The Australian Economic Review DOI: 10.1111/1467-8462.12312

°

C2018 The University of Melbourne, Melbourne Institute: Applied Economic & Social Research,

Faculty of Business and Economics

Published by John Wiley & Sons Australia, Ltd

The Australian Economic Review, vol. 52, no. 2, pp. 158–177

Hyslop and Townsend: The Longer-term Impacts of Job Displacement 159

We use linked data from two sources within

Statistics New Zealand’s Integrated Data Infra-

structure (IDI). The longitudinal Survey of

Family Income and Employment (SoFIE) is

used to identify workers who experienced a

job displacement between 2001 and 2010. The

SoFIE sample is thenmatched to administrative

data on monthly employment, earnings and

welfare benefits from April 1999 untilSeptem-

ber 2015. This provides a longer follow-up

period than available usingonly SoFIE, as well

as richer pre-displacement controls, and avoids

sample attrition from SoFIE. In particular, we

observe at least two and a half years before

and five years after any job displacement event

observed in SoFIE.

Our analysis uses complementary economet-

ric approaches to analyse the impacts of job

displacement on workers’employment, earn-

ings and receiptof welfare benefits. First, we use

regression methods to estimate impacts, con-

trolling for differences in workers’observable

socio-economic factors and fixed unobserved

effects.Second, following recent literature(e.g.,

Eliason and Storrie 2006), we select matched

comparison samples of non-displaced workers

with similar characteristics and labour market

histories to displaced workers,and estimate the

displacementimpacts by comparing the average

outcomes of the displaced and non-displaced

comparison workers.

To start, we present event studies that

compare the outcomes for displaced and non-

displaced workers, using alternative compari-

son groups. First, we select a random sample of

non-displaced workers employed in months

that displacements occurred. Second, to control

for differences between displaced and non-

displaced workers, we select a sample of non-

displaced workers whose characteristics and

pre-displacement employment and earnings

match those of the displaced workers. To

examine whether the impacts are associated

with ending a job, we also compare trends for

non-displaced workers who report a job end.

Each of these analyses suggest substantial and

persistent adverse impacts of displacement.

We then present the econometric analyses

using each of the comparison samples. The

results from these approaches are remarkably

similar. Displaced workers’employment rate

in the year following displacement was 20–25

percentage points (pp) lower than comparable

non-displaced workers and was still 8–12pp

lower after five years. Conditional on finding

employment, displaced workers’earnings were

25–30 per cent lower in the first year and 13–22

per cent lower five years after being displaced.

Although these adverse employment and

earnings effects are partly offset by benefit

receipt, such income support is modest, with

displaced workers’total income about 30 per

cent and 20 per cent lower one and five years

after displacement. The impacts are concen-

trated among older workers and are larger for

workers displaced during the recession, mainly

due to larger employment effects.

Our estimated effects are comparable to

existing estimates in other countries, suggest-

ing that large displacement costs cannot be

attributed solely to employment protection.

While job displacement is more costly in a

weak labour market, we find sizeable displace-

ment impacts even when the labour market is

tight. The impacts are substantially larger for

older workers, consistent with the notion that

they are due to the loss of firm- or industry-

specific skills.

The article is organised as follows. In the

next section we review the literature on the

impact of involuntary job loss. In Section 3

we describe the SoFIE and administrative

data used, the definition of job displacement,

and the sample characteristics. Section 4

presents the two approaches we use to

estimate the impacts of job displacement on

workers’subsequent outcomes. The article

concludes with a summary of the main results

in Section 5.

2. Literature Review

There are several reasons why involuntary job

loss may affect employment and wage oppor-

tunities. Theseinclude the loss of firm or sector-

specific human capital, job-match quality

between the worker and firm, and union or

industry wage premiums,which are expected to

be particularly strong for workers with greater

seniority and tenure in a firm. In addition,

°

C2018 The University of Melbourne, Melbourne Institute: Applied Economic & Social Research, Faculty of Business and

Economics