Bureaucratic Incentives and the Limits of New Public Management

A joke that turned out to be a theory

In November 1955, the naval historian C. Northcote Parkinson published a short, satirical essay in The Economist making an observation he intended, at least partly, as a joke: “work expands so as to fill the time available for its completion.” The line survives today as a piece of workplace folklore. Less well remembered is the evidence Parkinson actually offered for a second, more serious claim buried in the same essay — that the number of people working inside a bureaucracy grows at a fairly constant rate, regardless of how much work there is to do. His example was the British Colonial Office, whose staff had grown by roughly 5–7 per cent a year between 1935 and 1954, a period in which Britain's colonial territories were shrinking, not expanding. The Office reached its largest staff size at almost the exact moment it was folded into the Foreign Office, for lack of any colonies left to administer.

Parkinson meant this as comedy. Within two decades, an economist had turned the same intuition into a formal, testable theory of why public bureaucracies behave the way they do.

The theory: bureaucrats maximise budgets, not efficiency

William Niskanen's 1971 model of the “budget-maximising bureaucrat” provided the serious version of Parkinson's joke. In a private firm, Niskanen argued, the market imposes a fairly blunt discipline: sustained inefficiency threatens the firm's ability to pay its own bills, and firms that fail this test do not survive to keep failing it. A government bureau faces no equivalent test. Its budget is negotiated with a political sponsor rather than earned by satisfying paying customers in competition with rivals, and the official running it has every rational incentive to maximise the size and scope of the operation — more budget, more staff, more programmes — because nothing external punishes the alternative. Niskanen was careful, and it is worth being equally careful now, not to claim that no mechanism at all works against inefficiency in government: political accountability, audit, inspection and professional pride all pull in the other direction, to varying degrees. His claim was narrower and more defensible than that: the specific, blunt discipline a firm faces from its own survival is largely absent, and whatever does the equivalent job in government evidently does not do it with the same force.

The response: import the market, deliberately

If a bureau lacks market discipline naturally, one obvious fix is to import it artificially — and from the late 1970s onward, that is broadly what a wave of English-speaking governments set out to do, under the banner that became known as New Public Management (NPM). The idea took hold first and most visibly under Margaret Thatcher's government in Britain, but it was never a purely British phenomenon. It spread across the OECD through the 1980s and 1990s, in Australia, Sweden, the United States and elsewhere, with each country adapting the same basic instinct — competition, contracting-out, performance measurement, management by results — to its own institutions.

No country pursued the idea more completely than New Zealand. Between 1986 and 1989, three pieces of legislation — the State-Owned Enterprises Act, the State Sector Act and the Public Finance Act — restructured the New Zealand state so thoroughly that the public administration scholar Jonathan Boston has described the resulting model as widely regarded as “the purest embodiment of the principles and practices” of NPM anywhere in the world. State enterprises were corporatised and in many cases privatised; senior public managers were moved onto individual, performance-based contracts; and government accounting was shifted from a simple cash basis, common practice almost everywhere at the time, to full accrual accounting — a change significant enough that other countries' Treasury officials began making regular visits to Wellington to study it. New Zealand's reforms were pursued with unusual speed partly because its political system made that possible: a single-chamber Parliament, first-past-the-post elections producing decisive majorities, and no codified constitution slowing the process down. The reforms were also, notably, introduced by a Labour government responding to a genuine economic crisis — New Zealand's GNP per capita had fallen from fifth in the world in the 1950s to twentieth by the early 1980s — which is a useful corrective to any assumption that NPM was purely a project of the political right.

The United States pursued a parallel, somewhat less radical version of the same instinct through the 1990s, popularised by David Osborne and Ted Gaebler's influential 1992 book Reinventing Government and formalised in Vice President Al Gore's National Performance Review. The details differed by country, but the underlying diagnosis was the same one Niskanen had formalised two decades earlier: government does not face the discipline a market imposes, so government would have to impose something like it on itself.

The critique: efficiency is not the only value at stake

NPM acquired serious academic critics almost as quickly as it acquired governments willing to try it. Christopher Hood's foundational 1991 analysis, “A Public Management for All Seasons?”, argued that the doctrine's central and most persistent tension is between “efficiency” and “equity” values: public goods cannot simply be allocated by profit logic the way private goods can, and a relentless focus on efficiency risks displacing other things a market test was never designed to protect — reliability, fairness, continuity of care for people who cannot simply take their business elsewhere. Almost twenty-five years later, Hood and a co-author reviewed three decades of the reform in practice and titled their assessment plainly: higher costs, more complaints (Hood and Dixon, 2015) — a blunt verdict on a reform movement that had, by then, had ample time to prove its own case.

A different prediction: bureaucrats don't want bigger budgets, they want better jobs

Not every response to Niskanen accepted his starting assumption. The political scientist Patrick Dunleavy, in his 1991 book Democracy, Bureaucracy and Public Choice, proposed a direct alternative he called the “bureau-shaping model.” Dunleavy's central objection was that Niskanen's rational bureaucrat, single-mindedly chasing a bigger budget, does not match how senior officials actually seem to behave. A large budget also means more staff to manage, more routine administration, more scrutiny and more of the tedious work of running a big organisation. Dunleavy argued that what a rational, career-minded official actually wants is a small, elite, high-status agency doing interesting, high-level work — policy advice rather than delivery, strategy rather than casework — and that this preference, not budget-maximisation, better explains a pattern NPM reforms made increasingly common: senior civil servants who are often keen, not reluctant, to hive off the large-scale delivery functions of their own department into arm's-length agencies, contracted-out services, or privatised bodies. Where Niskanen predicts empire-building, Dunleavy predicts a preference for outsourcing and agencification — both models explaining the same underlying incentive problem, but pointing toward almost opposite behaviour.

What fifty years of this argument has actually settled

Very little of this is fully resolved, and that is itself the honest conclusion. Niskanen correctly identified that public bureaux lack the market's blunt discipline; Dunleavy correctly identified that the specific prediction Niskanen drew from that fact does not always match how officials actually behave in practice. NPM was a real, large-scale, multi-country attempt to solve the problem Niskanen described, tried with unusual thoroughness in New Zealand and unusual caution in the countries that adopted only fragments of it; and Hood's now three-decade-old warning that efficiency-first reform can quietly erode the other things public services exist to protect has, on Hood and Dixon's own later account, held up reasonably well.

What Parkinson noticed as a joke in 1955, in other words, turned out to describe a genuine structural problem that half a century of serious theory and several genuine national experiments have not fully solved, only clarified. Government bureaux do behave differently from firms, for identifiable reasons; importing market discipline changes that behaviour, but not without cost; and predicting exactly how a rational official will respond to the incentives they actually face has turned out to be considerably harder than either Niskanen or his critics initially assumed.

---

References

Dunleavy, P. (1991) Democracy, Bureaucracy and Public Choice: Economic Explanations in Political Science. New York/London: Harvester Wheatsheaf.

Hood, C. (1991) ‘A Public Management for All Seasons?’, Public Administration, 69(1), pp. 3–19.

Hood, C. and Dixon, R. (2015) ‘What We Have to Show for 30 Years of New Public Management: Higher Costs, More Complaints’, Governance, 28(3), pp. 265–267.

Niskanen, W.A. (1971) Bureaucracy and Representative Government. Chicago: Aldine-Atherton.

Osborne, D. and Gaebler, T. (1992) Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector. Reading, MA: Addison-Wesley.

Parkinson, C.N. (1955) ‘Parkinson's Law’, The Economist, November 1955.

Parkinson, C.N. (1958) Parkinson's Law: The Pursuit of Progress. London: John Murray.