Organisational form and individual motivation: public ownership, privatisation and fat cats

Abstract

We add potential intrinsic motivation to an agency model that is applied on public ownership and privatisation. Conventional agency theory suggests private ownership to be superior if pay under public ownership is not performance-related, but the ranking is otherwise reversed. However, we predict that motivation crowding out (MCO) can cause performance differences to go either way in both cases. Fat-cat salaries occur if public ownership with intrinsic motivation and a fixed wage is followed by privatisation with MCO, performance-related pay and a lower effort. The analysis also identifies factors that affect the performance of a given type of organisation.

Keywords:

• public and private provision
• intrinsic motivation
• motivation crowding out

JEL Classifications:

• L33
• H42
• D23
• M52

Acknowledgement

We thank Niklas Grönblom for research assistance, and participants in the CIRIEC-workshop Recent advances on public missions and performance – Theory contributions and literature review (Berlin, 2013), in the seminar organised by Aboa Centre for Economics (2013), and in our Departmental workshop (2012 and 2013) for valuable comments. We are also grateful for the suggestions from the editors and referees of this journal. Sonja Grönblom is grateful for financial assistance from the Victoria Foundation, The Foundation of Dagmar and Ferdinand Jacobsson, The Fabian Klingendahl Foundation, The Foundation of Agneta and Carl-Erik Olin, The Hans Bang Foundation, and the Foundation for Business Education in Turku.

Notes

1. There is evidence of a positive relationship between growth and the size of the state enterprise sector in the OECD countries (Fowler and Richards 1995). As for Finland, creating state enterprises has hardly been completely misguided, given that GDP per capita was 8.7 times higher in 1998 than in 1913, as compared to the Figures 4.6, 3.8, 6.0, 5.2 and 5.6 for Germany, UK, Sweden, US and China (Maddison 2001).

2. The figure was 18–22% before the privatisation wave in the 1990s, as compared to 14% in Austria, 11% in Britain and 6% in Sweden (Willner 2006).

3. Typically, one of the early academic contributions, Kay and Thompson (1986), was titled “Privatisation: A Policy in Search of a Rationale”.

4. Wider objectives are also believed to produce complicated chains of command, but this can depend on details of implementation rather than ownership per se. Moreover, post-privatisation regulation might lead to similar complexity.

5. Also, the main culprit is democracy and not state ownership as such if decisions are distorted because of a desire to please voters. This would suggest that state ownership is more efficient under dictatorship (such as in the former Soviet bloc) than under democracy, which is scarcely convincing.

6. This was true in the study of the period 1974–1986 by Jensen and Murphy (1990). However, a widespread view that CEOs were under-incentivised caused payments to become highly sensitive to changes in shareholder value, which has led to new agency problems related to earnings manipulation (Bergstresser and Philippon 2006).

7. For example, Bénabou and Tirole (2003) refer to how Tom Sawyer gets his friends to compensate him for getting the perceived privilege to paint a fence.

8. Grönblom and Willner (2008) deal with public and private ownership in the presence of excess wages that are included in the total surplus.

9. The privatisation of refuse collection forced households in some British cities to roll their dustbins from their back-gardens to the pavement before being emptied, and to roll them back again afterwards. This may have saved money, but partly by forcing customers to do part of the job.

10. To assume the slope −1 is equivalent to normalising all prices by dividing them by the original slope of the inverse demand function.

11. The number of firms is given and fairly small because of the plausible assumption of sluggish entry (see Geroski 1995), and because perfect competition is rarely a feasible alternative to a public monopoly. But in this section and in Appendix 2 we also address free entry.

12. Strictly speaking, this assumption implies an infinite range of u, but it is a convenient simplification also if u belongs to a finite interval and if the distribution is bell-shaped and such that.

13. For the conditions under which a linear compensation scheme is optimal, see Holmstrom and Milgrom (1987).

14. It would be possible to assume v₀ = 0 in this section (like in Beiner, Schmid, and Wanzenried 2011) but not in Section 4, where the assumption might lead to negative wages.

15. Note that a weighted objective function of the type αE(CS)+(1 − α)E(π) yields the same solutions as (3.7), as can be seen by dividing by 1 − α, and denoting α/(1 − α) by ρ. Note, however, that α = 1 would require a break-even constraint or a maximum subsidy, because E(CS) is non-concave (see also Appendix 2). We might alternatively use a weight for the total surplus (because ρE(CS) + E(π) = ρ(E(CS) + E(π))+(1 − ρ)E(π)), or for output (see Willner 2013).

16. Normalising the slope of inverse demand to −1 implies that CS is the triangle (a − p)x/2 = x²/2.

17. This requires the stringent but widely used assumption that random shocks in different firms are independent, and hence a focus on firm-specific risks only. Common risks are assumed to be observable.

18. Willner and Grönblom (2009) and Sugden and Valania (2013) adopt a different approach, where the agent is partly opportunistic, partly committed. She decides according to a cooperative intra-personal Nash-game, in which PRP can cause opportunism to dominate.

19. Apart from some symbols, the only difference as compared to James (2005) is that we have replaced by .

20. Goal identification would mean replacing by a term that relates, for example, to the organisation’s output (see Appendix 2).

21. The employer may be ignorant about the significance of for MCO, for example because of managerial fashions that are inspired by public choice theory or simple agency models that ignore intrinsic motivation. (On the economic thinking behind the New Public Management (NPM), see Gruening 2001; on NPM and its performance targets, see Hood 1995.)

22. A welfare and profit maximising monopoly would yield the TS_G= (a − c_G)²/2 and TS_P= 3(a − c_P)²/8, respectively; so c_G> c_P is a necessary but not sufficient condition for TS_P > TS_G.

23. The upper limits for breaking even for and n are decreasing in and , because they work technically in the same ways as sunk costs.

24. The NPM has meant replacing informal norms by explicit and measurable standards of performance (Hood 1995), which would cause MCO if they make it impossible to derive satisfaction from overperforming.