Control patterns in contracting-out relationships: it matters what you do, not who you are
Summary (2 min read)
Introduction
- They find that, consistently with literature expectations, the delivery mode (in-house, delivery by government, by for-profit or nonprofit entities) can actually affect the intensity of the controls put in place as a consequence of the divergence in the provider’s goals.
- Control patterns tend to combine and, while environmental (uncertainty and asset specificity) and task ( output measurability and task programmability) characteristics only partially explain the adoption of certain control configurations, control mechanisms seem to be influenced by party characteristics (partner knowledge, ownership configuration and political visibility).
- Other variables might aid explanation, as Hefetz and Warner (2012) show for contracting-out, and Cristofoli et al. (2011) suggest for control decisions.
- Previous contracting-out literature has identified the nature of the service provider (forprofit, non-profit, public) as a possible explanatory factor of the intensity of controls (Marvel and Marvel 2007, 2008; Van Slyke 2007).
Methods
- In order to carry out the research the authors conducted a survey for all Italian municipalities (510) of at least 20.000 inhabitants.
- These services were chosen in order to ensure diversity in the variables under consideration (Brown and Potoski 2003)2.
- The authors obtained a total of 91 responses3, where 46% referred to solid waste collection and 54% to homecare for elderly.
- To explore the proposed associations, the authors ran three separate linear regressions where against each type of control mechanism (market-based, hierarchy-based, trust-based), they regressed the independent variables previously discussed.
- Old contract length+β8Turnover+β9Political orientationi+β9 Population +β10 Geographical areai +β11 Surplus/deficit +ɛ.
Control patterns
- 3 In order to account for a non-response bias, the authors ran a two-sample t-test, using late respondents as surrogates for nonrespondents (Wallace and Mellor, 1988).
- The difference across responses was not statistically significant.
- - 13 - Considering the descriptive statistics presented in Table 2, control patterns’ mix and minimum and maximum scores relative to the three control types suggest that the trustbased mechanisms are the most present, followed by hierarchy- and market-based ones.
- Looking across services, while trust controls are still predominant (Table 3), hierarchybased controls are more present than market-based ones in the elderly homecare; vice-versa for waste collection.
Service and relationship characteristics
- In order to explore the variables associated with the different types of control (i.e. market-, hierarchy- and trust-based), three separate sets of linear regressions were performed (Tables 5, 6 and 7).
- As expected, the significance of the explanatory variables varies with the type of control adopted.
- Proposition 1 appears to hold for hierarchical and market-based patterns of control, but not for trust-based ones (Table 4).
- Interestingly, service characteristics appear to be significantly associated with the two former types of control, but not with the latter.
- Only political visibility appears to be significantly (at 1%) related to market-based controls.
Insert Table 4
- Explaining control mechanisms in contracting-out - 14 - Looking more closely at the explanatory variables, market-based mechanisms are positively associated with political visibility (also representing the different types of services) and output measurability (both significant at 1%), while negatively related to task uncertainty (at 5% significance).
- With respect to political visibility, the data show that waste collection services, which are the most visible to citizens, foster market-type controls.
- This finding contradicts their expectations and previous studies (Lioukas et al. 1993; Cristofoli et al. 2010).
- Hierarchy-based mechanisms are more clearly explained by the identified variables, with an R2 of 54.37% defined by the significant association with asset specificity (10%) and output measurability (1%), and the negative relationship with the municipality’s financial performance (significant at 5%).
- It is also worth noting that service political visibility does not contribute to an explanation regarding the adoption of hierarchy-based controls.
Insert Tables 5 to 7
- Control mechanisms and explanatory variables: an interpretation.
- On the contrary, once trust and hierarchical bases have been laid, the results suggest that market controls also play a role, and selective choices are made when these mechanisms are deemed useful.
- This leads to a triangulation of different mechanisms (in addition to the more common trust- and hierarchy-based ones) to monitor the service and its provider.
- This attitude - 18 - might bring about a relative weakness in control systems’ design, which, in turn, may also account for the low variance across the different delivery modes (majority of non-profits in elderly homecare, majority of public-sector providers in waste collection).
- This confirms that market-based controls are not necessarily tied to the mode of governance adopted, and therefore control practices can be of a different nature to the organizational forms chosen for public-service provision (Spekle 2001; Caglio and Ditillo 2008a).
Conclusions
- Public-service contracting-out has often been accompanied by a strong academic focus on the emergence of new governance forms, and a major neglect of the processes and practices through which these contracted-out services are controlled and monitored.
- In particular, bringing together variables so far investigated separately in the inter-organizational control and contracting-out literatures, the authors - 19 - explored whether the variables traditionally used in these studies have explanatory potential with reference to control choices of contracted-out services.
- Trust-based controls, in particular, are the most widespread, but cannot be explained by the contractingout and inter-organizational variables traditionally used.
- Finally, different from expectations (Marvel and Marvel 2007, 2008), mode of delivery and goal congruence do not appear to influence significantly the type of control adopted.
Did you find this useful? Give us your feedback
Citations
114 citations
Cites background from "Control patterns in contracting-out..."
...First, studies drawing on functionalist approaches have looked at the antecedents of control systems in hybridized and contracted out services (eg., Cristofoli et al., 2011; Ditillo et al., 2015; Johansson et al., 2016)....
[...]
...However, less attention has been devoted to accounting systems in such contexts from a public administration perspective (see, Ditillo et al., 2015)....
[...]
51 citations
50 citations
41 citations
Cites background from "Control patterns in contracting-out..."
...We should note that state owned enterprises were included in CRIPO’s ‘agency typology’ (Verhoest et al. 2012) and we share the idea that a variety of control mechanisms exists within the same governance structure (Ditillo et al. 2015)....
[...]
...When services display moderate asset specificity, task interdependence, and output measurability, private sector experience emphasizes the adoption of process control (Ditillo et al. 2015)....
[...]
40 citations
References
5,819 citations
"Control patterns in contracting-out..." refers background in this paper
...Only a few studies, drawing mainly on economic theories, such as transaction-cost economics (Williamson 1981) and agency theory (Jensen and Meckling 1996), have focused on the monitoring of public services in contracting-out contexts....
[...]
3,262 citations
"Control patterns in contracting-out..." refers background in this paper
...Finally, trustbased patterns are generally described as substitutes or complements for formal control mechanisms (Dekker 2004; Das and Teng 1998), based on social controls, broad nonspecific contracts and emergent performance assessment standards, personal consultation and intensive communication…...
[...]
1,705 citations
"Control patterns in contracting-out..." refers methods in this paper
...This variable has been alternatively called delivery mode (Marvel and Marvel 2007, 2008) or goal congruence (Van Slyke 2007; Provan and Milward 1995)....
[...]
1,026 citations
"Control patterns in contracting-out..." refers background in this paper
...Finally, a few studies suggest that higher knowledge of the partner can result in either a reduced need for formal coordination or improved coordination (Dekker 2004; Cristofoli et al 2010)....
[...]
...Finally, trustbased patterns are generally described as substitutes or complements for formal control mechanisms (Dekker 2004; Das and Teng 1998), based on social controls, broad nonspecific contracts and emergent performance assessment standards, personal consultation and intensive communication…...
[...]
...Finally, trustbased patterns are generally described as substitutes or complements for formal control mechanisms (Dekker 2004; Das and Teng 1998), based on social controls, broad nonspecific contracts and emergent performance assessment standards, personal consultation and intensive communication to generate confidence that the other parties will not behave opportunistically (Langfield-Smith and Smith 2003)....
[...]
1,018 citations
Related Papers (5)
Frequently Asked Questions (11)
Q2. What have the authors stated for future works in "Control patterns in contracting-out relationships: it matters what you do, not who you are" ?
First, the authors assessed the explanatory potential of transaction-cost variables with reference to control choices, rather than to contracting-out choices as most of the literature had done before. The study calls for further research on services with different characteristics. Future studies might devote more attention towards the political and contextual factors affecting publicservice management ( Pollitt 2013 ) 4, which can not simply be grasped through quantitative analysis. The importance of trust-based controls and the difficulty in explaining their adoption with respect to traditional variables ( mainly drawn on economic and control studies ) suggest that additional studies are needed to explore the role of other variables, such as political and informal factors, in defining the type and the intensity of the controls used between the - 20 - public sector and public-service providers.
Q3. What is the need for the development of managerial competences and skills?
The current pervasiveness of trust-based mechanisms also points to the need for the development of managerial competences and skills, such as communication, negotiation and networking, together with the strengthening of transparency mechanisms in place between the public sector and service providers.
Q4. What is the main reason for the change in the public sector?
as a consequence of the managerial reforms that took place over the last two decades in the public sector, a greater emphasis has been put on the communication of results and output as a means of providing legitimation for public sector activities.
Q5. What is the important factor in determining the intensity of controls?
high degrees of alignment between the public sector’s and the provider’s objectives (i.e. goal congruence) should result in less intense controls.
Q6. How did the model account for a non-response bias?
In order to account for a non-response bias, the authors ran a two-sample t-test, using late respondents as surrogates for nonrespondents (Wallace and Mellor, 1988).
Q7. What are the main reasons for the lack of monitoring systems?
According to these, monitoring systems are put in place in order to minimize transaction costs and goal misalignment between principal and agent.
Q8. Why do more politically visible services need tighter controls?
more politically visible services (such as waste collection) might require tighter overall controls because of their greater importance.
Q9. What methods of research are needed to study informal processes of control?
Consistent with this view, while the authors conducted a survey, studying informal processes of control might require more participative research methods, such as interviews, direct observations and ethnographic studies.
Q10. What is the main reason for the preference for hierarchical mechanisms?
The preference for hierarchical mechanisms is probably further strengthened by the traditional bureaucratic culture of the country under analysis, where the accomplishment of predefined rules is considered as the achievement of the result itself (Cristofoli et al. 2010).
Q11. What are the main variables that may be relevant in explaining contracting-out choices?
The characteristics of the relationship (such as mode of delivery/goal congruence, political visibility and partner knowledge) between the parties involved in the public-service provision may be relevant not only in explaining contracting-out choices, but also in defining the types of controls subsequently set in place.