This essay formed part of my MSc in International Public Policy. I’ve rewritten it here to be easier to read, but the argument and sources are the same.
This is the first of my posts purely about public policy. It looks at the last stage of making policy: implementation, the point where a decision on paper has to become something real. It’s arguably the most important stage, and it’s where a lot of good ideas go wrong. Below I explain why it’s so hard, using three real cases.
In short
- How you define implementation shapes how you go about it, especially for the people on the front line.
- Policies designed only from the top down tend to ignore the people who actually deliver them.
- History (path dependence), self-interest and discretion show up in every case.
- Unintended outcomes are guaranteed, so plan for them.
- No single approach works everywhere. The best results come from balancing them.
Why implementation matters
Today everyone accepts that implementation is important and complicated, but that hasn’t always been true. In the early 1960s, policy analysis focused almost entirely on the ‘front end’ of the process: deciding what to do. Implementation was treated as a background detail. The hope was that by studying where policies failed at the delivery stage, policymakers could learn from those failures.
This essay asks why implementation is so difficult. My main case study is the Economic Development Administration (EDA) project in Oakland, California, with further cases from Nigeria and the UK. I’ll cover:
- three different definitions of implementation, and why the definition matters at ‘street level’, where front-line staff use their own judgement;
- the strengths and weaknesses of the top-down and bottom-up approaches;
- how far institutions and policies are trapped by their own history (path dependence);
- why progress is hard to measure, what is outside a decision maker’s control, and what New Public Management (NPM) offers;
- some practical lessons for making implementation work.
This isn’t an exhaustive list. Policies fail for many different reasons, and no two cases are the same.
What is implementation?
At its simplest, implementation means turning policy into practice. But how an official goes about it depends a lot on how they define it. Here are three well-known definitions:
- Pressman and Wildavsky: implementation means ‘to carry out, accomplish, fulfil, produce, complete’ (Pressman and Wildavsky, 1984). They also ask a good question: where does a policy end and its implementation begin? Their definition stresses the job of public servants to deliver efficient, fair services to citizens.
- Mazmanian and Sabatier (top-down): implementation is ‘carrying out a basic policy decision, usually incorporated in a statute but which can also take the form of important executive orders or court decisions’ (Mazmanian and Sabatier, 1983).
- Van Meter and Van Horn (bottom-up): implementation is ‘those actions by public or private individuals (or groups) that are directed at the achievement of objectives set forth in prior policy decisions’ (Van Meter and Van Horn, 1975).
The top-down view and the ‘implementation gap’
Because early policy analysis concentrated on the top end and neglected the ‘rear end’, it ran into problems. Some were about method. Others were about researchers taking on the views of the people they studied about what should happen (Hill and Varone, 2017).
This ‘first wave’ of implementation research focused only on the gap between what was planned and what was delivered, and it was only partly successful (deLeon and deLeon, 2002). These early studies tended to blame failure on badly written laws or officials not following orders (Schofield, 2001).
The top-down approach rests on two assumptions (Cairney, 2012):
- Descriptive: decisions are made at the top (central government or parliament) and carried out at the bottom (the organisations delivering them).
- Prescriptive: decisions should be made at the top and carried out at the bottom.
Both top-down and bottom-up approaches left big implementation gaps: the difference between what policymakers expected and what actually happened (Cairney, 2012). Those gaps pushed researchers to dig deeper, and by the 1990s many were combining the two approaches.
Case study 1: the Oakland project
In their landmark study from the early 1970s, Pressman and Wildavsky assumed policy moved in one direction: made at the top, then carried out by officials. Even so, they could see that implementing the EDA’s plans in Oakland would be difficult from the start.
The EDA had a single goal: reduce unemployment. But delivering it meant following two separate paths: paying for construction work, and creating a hiring plan that included African-American workers. Along each path, unexpected decisions kept appearing and the number of approvals needed kept growing (Pressman and Wildavsky, 1973). None of this had been in the original plan.
Even if both paths had been planned from day one, it’s not clear the project could have been saved. The people at the bottom were never considered, and little thought was given to unintended consequences. The workers who were consulted would also become decision makers themselves, using their own judgement. Browne and Wildavsky (1983) later recognised that many causes interact, but discretion wasn’t properly studied in their work at the time.
Case study 2: education policy in Nigeria
The second case is Nigeria’s National Policy on Education, created in 1977. As a developing country, Nigeria is a sharp contrast to the other cases here. The policy aimed to give people the skills, both mental and physical, to live in and contribute to Nigerian society (Ikechukwu Ugwuanyi and Chukwuemeka, 2013).
Nigeria has made big progress in getting children into primary school, but UNESCO concluded that more still needs to be done on both numbers and quality (Irigoyen, 2017).
Weak planning and poor communication
The 1977 policy was designed almost entirely from the top, shaped by previous education policy and culture, and that pattern has continued. UNESCO found that the curriculum is undermined by the generally low quality of the teachers who deliver it, which leads to low achievement (Irigoyen, 2017).
As in Oakland, involvement of the people affected was weak. There was no sign of consultation between federal and state institutions when the Universal Basic Education programme was designed, which caused misalignment later (Irigoyen, 2017). That poor communication also caused problems with how money was shared between states. Again, there was a big gap between what was planned and what was delivered.
A country’s location, culture, politics, economy and society all have to be taken into account, especially at the implementation stage. That’s why comparing developing and developed countries directly isn’t really fair.
Not enough people, money or say for teachers
The policy leaned heavily on state agencies that didn’t have the staff or money to deliver it. Teachers were the ones putting it into practice, yet nobody thought to involve them at the start. Money was part of the problem too: the national government sometimes doesn’t budget enough for officials to implement policies properly (Ikechukwu Ugwuanyi and Chukwuemeka, 2013).
Stuck on an old path
Part of the reason for the lack of bottom-up thinking may be how the Nigerian government defines implementation, and how far its institutions are tied to the past. Nigeria only gained independence from Britain in 1960, and building its own education system took time amid political and economic challenges. Before 1977 it used an education policy inherited from Britain (Irigoyen, 2017). Under colonial rule, policy was fixed, with very little input from the people delivering it.
This is path dependence: once a commitment is made and resources are put behind it, it becomes more and more costly to change direction (Cairney, 2012, citing Pierson, 2000). In Nigeria, policymakers have often been unwilling or unable to change course.
Unclear rules at street level
By focusing on the top, the government didn’t allow for the discretion and uncertainty faced by ‘street-level’ staff such as teachers and head teachers. The policy is very vague about how they should actually deliver it, with no clear management structures or guidelines (Irigoyen, 2017).
Keeping personal interest, prejudice and old habits out of official business is also a serious challenge in Nigeria (Ikechukwu Ugwuanyi and Chukwuemeka, 2013). Each new government tends to want to put its own stamp on public programmes, and corruption and weak leadership damage the quality of policy before it’s even implemented. If a policy is botched at the design stage, it’s unlikely to succeed however involved the people on the ground are.
Bottom-up theorists argue that if local staff aren’t free to adapt a policy to local conditions, it will probably fail (Matland, 1995). I’d say that depends on the case. That freedom only really works when front-line staff are fully involved and the rules are clear. Teachers in Nigeria face two big problems they may have had to tackle on their own: social and cultural barriers to girls’ participation, and weak enforcement of the 2004 Universal Basic Education Act on enrolment and keeping children in school (Irigoyen, 2017).
A lack of basic training, communication and coordination with teachers led to low achievement, and to staff using a lot of discretion to make up for it with little oversight. Clear rules are essential here. It’s also worth remembering that front-line staff face the same limits as any decision maker: they can’t know everything (what’s called bounded rationality).
Did the 1998 changes help?
The 1998 edition of the policy set out specific roles for policymakers and implementers, and recommended creating a State Agency for Mass Education (SAME) (Irigoyen, 2017). This was meant to reduce confusion at street level. But because of the culture and history inside schools and other institutions, more government oversight and clearer rules are probably the only ways to fix the original failure.
It shows how path dependence makes later fixes harder: history matters, especially when a policy starts out vague. National institutions are still poorly aligned, with overlapping organisations and unclear goals (Irigoyen, 2017), so confusion remains high.
Self-interest: who gains?
Self-interest is present at every stage. Politicians may want to be seen to support a cause while doing nothing about it (Hill and Varone, 2017). Both Oakland and Nigeria show this.
The interests of the principal (elected officials and institutions) and the agent (the official who has to deliver) can pull in different directions. This was true in Oakland and in UK rail privatisation, where British Rail managers used their networks to win extra resources. After objecting to the form privatisation took, these managers went on to collaborate fully, and some became wealthy as a result (Grantham, 2001). Where money is at stake, people manoeuvre to secure it for themselves (Grantham, 2001). On paper the managers had little discretion, but their financial knowledge gave them real power (Grantham, 2001).
Case study 3: UK rail privatisation and New Public Management
Later research moved away from the extremes. Saetren (2014) describes ‘the gradual silencing of the protracted and unproductive debate between top-down and bottom-up scholars’. UK rail privatisation fits this later view: it’s more of a balance between the two.
New Public Management (NPM) set out to close the implementation gaps that Oakland and Nigeria illustrate. It rose to prominence in the early 1980s under Margaret Thatcher in the UK and Ronald Reagan in the US. The idea was that the central state was ‘overloaded’: monitoring, oversight and performance targets needed to increase, and the discretion of public sector professionals needed to shrink. Internal markets and fixed-term contracts were central, justified by ‘rivalry as the key to lower costs and better standards; contracts as the key to explicating performance standards’ (Hood, 1995).
Traditional top-down and bottom-up approaches only explain part of what happens. NPM tried to measure and monitor outcomes (Grantham, 2001), solving a problem the earlier approaches had largely ignored.
For Grantham, the success of rail privatisation depended mainly on activity in a series of separate implementation networks that existed only during the move to privatisation (Grantham, 2001). NPM was visible throughout, with grants and direct payments from the Treasury. Competitive contracts for private train operators were expected to cut costs and improve the service at the same time.
Supporters say performance indicators, targets and new technology make implementation more effective, stressing the ‘need to cut direct costs, raise labour discipline, and to do more with less’ (Hood, 1995). Critics say NPM treats public servants as wasteful and has demoralised the public sector. It remains a deeply contested approach, for reasons beyond the scope of this essay.
Expect the unexpected
One theme runs through all three cases: unintended outcomes. In fact, it runs through implementation as a whole. As former US Defense Secretary Donald Rumsfeld put it:
We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones.
Federal News Service Inc., 2002
The task for policymakers is to anticipate and prepare for unintended outcomes early on. Learning from similar past cases helps, as does setting aside money and staff for the unexpected. For example, Nigeria’s later focus (from 2004) on secondary education and workplace skills led to an unintended 4% fall in primary school enrolment (Irigoyen, 2017).
Why do policies fail?
It’s often hard to tell whether a policy failed because it was badly designed, badly delivered or a mix of both (Hill and Varone, 2017). Hogwood and Gunn (1984) point to three causes:
- bad execution;
- bad policy;
- bad luck.
Cairney (2012) adds that the implementation gap is wide because people have unrealistic expectations. All of these are fair points, but in practice the reasons for failure vary a lot from case to case.
Conclusion
The distance between the people who design a policy and the people who deliver it is critical to success. Policy is a complex, contested area, which makes good design harder, and that feeds straight into implementation. Politics, culture, economics and location all shape the result, so outcomes are unpredictable even in stable, developed countries.
This essay only covers three cases, each with different political incentives, goals and expectations. Most implementation research focuses on Western countries, so I included Nigeria to help balance that. The Nigerian case shows how much history, place and culture determine whether a policy meets its goals.
Some lessons stand out:
- Mind the gap: involve the people who will deliver a policy from the start.
- Be clear: vague policies leave front-line staff guessing.
- Know your history: path dependence makes change slow and costly.
- Allow for self-interest and discretion: they appear in every case.
- Plan for the unexpected: set aside resources for unintended outcomes.
- Balance approaches: NPM brings useful performance measurement but has real downsides. Learning from all perspectives works best.
Implementation is full of pitfalls, but they can be reduced.
References
Cairney, P. (2012) Understanding Public Policy, First. Hampshire, Palgrave Macmillan [Online]. (Accessed 25 October 2017).
deLeon, P. and deLeon, L. (2002) ‘What Ever Happened to Policy Implementation? An Alternative Approach’, Journal of Public Administration Research and Theory: J-PART, vol. 12, no. 4, pp. 467–492.
Federal News Service Inc. (2002) Defense.gov Transcript: DoD News Briefing – Secretary Rumsfeld and Gen. Myers [Online]. Available at http://archive.defense.gov/Transcripts/Transcript.aspx?TranscriptID=2636 (Accessed 27 February 2018).
Grantham, A. (2001) ‘How Networks Explain Unintended Policy Implementation Outcomes: The Case of UK Rail Privatization’, Public Administration, vol. 79, no. 4, pp. 851–870 [Online]. DOI: 10.1111/1467-9299.00283.
Hill, M. J. and Varone, F. (2017) The Public Policy Process, Seventh. London, New York, NY, Routledge, Taylor & Francis Group.
Hood, C. (1995) ‘Contemporary public management: a new global paradigm?’, Public Policy and Administration, vol. 10, no. 2, pp. 104–117 [Online]. DOI: 10.1177/095207679501000208.
Hogwood, B. and Gunn, L. (1984) Policy Analysis for the Real World (Oxford: Oxford University Press).
Ikechukwu Ugwuanyi, B. and Chukwuemeka (2013) ‘The Obstacles to Effective Policy Implementation by the Public Bureaucracy in Developing Nations: The Case of Nigeria’, Kuwait Chapter of Arabian Journal of Business and Management Review, vol. 2, pp. 59–68 [Online]. DOI: 10.12816/0001218.
Irigoyen, C. (2017) ‘Universal Basic Education in Nigeria’, Centre for Public Impact [Online]. Available at https://www.centreforpublicimpact.org/case-study/universal-basic-education-nigeria/ (Accessed 27 February 2018).
Matland, R. E. (1995) ‘Synthesizing the Implementation Literature: The Ambiguity-Conflict Model of Policy Implementation’, Journal of Public Administration Research and Theory: J-PART, vol. 5, no. 2, pp. 145–174.
Mazmanian, D.A. and Sabatier, P. (eds.) (1981) Effective Policy Implementation (Lexington: Lexington Books).
Meter, D. S. V. and Horn, C. E. V. (1975) ‘The Policy Implementation Process: A Conceptual Framework’, Administration & Society, vol. 6, no. 4, pp. 445–488 [Online]. DOI: 10.1177/009539977500600404.
Pressman, J. L. and Wildavsky, A. (2018) Implementation – How Great Expectations in Washington Are Dashed in Oakland; Or, Why It’s Amazing that Federal Programs Work at All, This Being a Saga of the Economic Development Administration as Told by Two Sympathetic Observers Who Seek to Build Morals on a Foundation of Ruined Hopes – The Oakland Project.
Saetren, H. (2014) ‘Implementing the third generation research paradigm in policy implementation research: An empirical assessment’, Public Policy and Administration, vol. 29, no. 2, pp. 84–105 [Online]. DOI: 10.1177/0952076713513487.
Schofield, J. (2001) ‘Time for a revival? Public policy implementation: a review of the literature and an agenda for future research’, International Journal of Management Reviews, vol. 3 (3), pp. 245–263.
