Published date01 September 2016
Date01 September 2016
doi: 10.1111/padm.12262
It is often said that ‘necessity is the mother of invention’. In this article, we assess whether this also
applies to the design of public authorities’ service provision. In particular, we evaluate whether
revenue scarcity – as an indicator of scal stress – induces government outsourcing. In contrast to
previous studies, we exploit arguably exogenous variation in local government revenue across time
and space to derive stronger inferences on the role of revenue scarcity for outsourcing. Using data
from Norwegian local governments covering the period 1995–2012, our main results indicate that
a decrease in local government revenues is linked to more outsourcing of both infrastructure and
support services.
An enduring debate in public administration concerns the institutional design of (local)
government service provision. In principle, several options are available to public author-
ities – ranging from in-house provision, over contracting out to public or private sector
providers (i.e. outsourcing) to full-scale privatization. This range of alternatives raises two
important questions. First, which form of service provision provides the ‘best’ outcomes
in terms of, for instance, cost or production-technical efciency, service effectiveness and
service quality (for an early review,see Domberger and Jensen 1997)? Second, what deter-
mines who chooses which form of service provision (for an early review,see Boyne 1998)?
Our study concentrates on the latter question, and thus aims to help elucidate why certain
jurisdictions turn to outsourcing, while others do not.
More specically, we concentrate on the potential role of revenue scarcity – or scal
stress – for the decision to outsource service provision. Early observers often view scal
stress as a key driver behind outsourcing and privatization decisions (Moore 1987; Boyne
1998). It is currently gaining increased attention again due to the scal crisis in many West-
ern countries, which ‘intensies the need for local governments to rethink service delivery’
(Bel and Warner2015, p. 52; see also Ceipku et al. 2016). Several theoretical arguments have
been brought forward for such a potential revenue–outsourcing relation. A rst line of
argument focuses on a simple production cost argument: afuent governments can offer
more and better in-house service provision compared to low-revenueauthorities, and thus
are less likely to rely – or, possibly, be forced to rely – on alternative provision systems
(Moore 1987). Conversely, governments suffering ‘nancial strains due to softening rev-
enues’ may be particularly likely to ‘feel pressure to cut back and possibly jettison noncore
functions as a way to deal with the problem’ (Lamothe and Lamothe 2016, p. 360).
Another line of argument focuses on political costs and benets, and highlights that
outsourcing might offer increased governance exibility by acting as a ‘budgetary buffer’
(Pallesen 2006, p. 39). When service provision is contracted out, the political and admin-
istrative costs of adjusting service levels are smaller. In line with such reasoning, recent
evidence suggests that local governments are more likely to shed services that have been
Benny Geys and Rune J. Sørensen are at the Norwegian Business School (BI), Oslo, Norway.
Public Administration Vol.94, No. 3, 2016 (769–788)
© 2016 John Wiley & Sons Ltd.
outsourced to external suppliers in the recent past (Lamothe and Lamothe 2016). Finally,
a third line of argument builds on a transaction cost perspective (Williamson 1981, 1996),
which maintains that public service providers will choose governance structures minimiz-
ing the transaction costs (e.g. costs of negotiating and contracting) involved in obtaining
a desired service level. In such a framework, cost pressures make it more likely that the
value assigned to a reduction in short-term costs via outsourcing will outweigh concerns
about any contracting costs linked to outsourcing (Williamson 1996; Klaas et al. 1999).
Overall, therefore, nancial difculties faced by local governments may increase the
appeal of outsourcing and privatization options. Yet, while indicators of scal stress are
among the most frequently considered variables in outsourcing or privatization studies
(Ferris 1986; Warner and Hebdon 2001; Bakker 2002; Pallesen 2004; Brudney et al. 2005;
Bhatti et al. 2009; Zullo 2009; Gonzalez-Gomez et al. 2011; Rho 2013), empirical results in
this literature have thus far remained inconclusive at best (for reviews, see Boyne 1998; Bel
and Fageda 2007, 2009). Moreover,previous work has been criticized for failing to account
for the ‘reciprocal relationships between contracting out and the explanatory variables’
(Boyne 1998, p. 150) and the ‘dynamic nature’ of the outsourcing/privatization decision
(Bel and Fageda 2007, p. 517). As such, inferences drawn from existing empirical studies
remain exceedingly fragile, and at best reect a (possibly spurious) correlation between
scal stress and outsourcing rather than a causal connection.
Our main contributions relate to tackling these two empirical deciencies in previous
studies. First, we have access to unique panel data, which cover six surveys on the institu-
tional design of local public authorities’ service provision collected in consistent four-year
intervals among Norwegian municipalities over the period 1996–2012 (with an additional
initial survey elded in 1995). Hence, we can study changes over time and space in rev-
enue scarcity and outsourcing to more accurately capture the move frompublic to private,
or vice versa. It should be noted here that several recent studies have likewise made use of
panel data (Pallesen 2004; Bhatti et al. 2009; Zullo 2009; Gonzalez-Gomez et al. 2011; Hefetz
and Warner2012; Sundell and Lapuente 2012; Rho 2013). Nonetheless, these often still rely
on pooled OLS regressions that effectively exploit only the cross-sectional variation in the
data, and thus ignore the information provided by any existing temporal variation (e.g.
Pallesen 2004; Bhatti et al. 2009; Sundell and Lapuente 2012).
Second, and crucially, our analysis can exploit variation in local government revenue
scarcity outside the direct control of local governments to provide a more credible identi-
cation of the true revenue–outsourcing relation. This is important because, as mentioned,
any decision to engage in outsourcing naturally also affects the level of government expen-
ditures and, therefore, revenue requirements (Boyne 1998). Ignoring this joint determina-
tion of revenues and outsourcing generates spurious results in the sense that the estimated
correlation between both variables is likely to be inconsistent (i.e. it will not reect the true
population parameter).
In contrast, and following Hægeland et al. (2012), Andersen et al. (2014) and Borge et al.
(2015), our approach builds on the fact that many Norwegian municipalities derive sig-
nicant revenue streams from the presence of hydropower plants within their bound-
aries. These revenues – which can amount to a substantial share of the local budget – are
independent of local government decision-making since all hydropower stations were
constructed in the period 1950–90 (i.e. prior to the period analysed here), their location
depends on the presence of mountainous terrain (which is beyond municipal control),
and municipalities in practice never adjust the commercial property tax rate levied on
hydropower plants (i.e. all set the maximum rate allowed by the national government)
Public Administration Vol.94, No. 3, 2016 (769–788)
© 2016 John Wiley& Sons Ltd.

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