Making good public-policy can be highly counter-intuitive, for a number of reasons. In this post I will propose four principles which I believe have the potential to improve the quality of many policies.
1. Respect street-level bureaucrats
I have discussed street-level bureaucracy in other places, because I think it is an extremely important part of why some policies do or do not succeed. As a refresher, street-level bureaucracy points to the phenomenon of how some public servants (e.g. teachers, police, nurses) define public policy through their daily behaviours. Street-level bureaucracy does not encompass all civil service workers, but rather specifically points to poorly resourced front-line workers who must translate public and political expectations into action serving the public. The street-level bureaucrats’ response to these pressures is to develop coping mechanisms to deal with the overwhelming workload. These include such measures as mentally dehumanizing their clients, establishing elaborate processes to mitigate and reduce the use of public services, adhering rigidly to routine, and many others.
The key thing that policy makers need to attend to in regards to street-level bureaucracy is the importance of remembering and respecting the civil-service workers that enact their policies. I remember speaking to a friend on how the police subvert and manipulate crime statistics to game the system, by which I was arguing that centralized targets can do more harm than good. My friend’s response was to suggest that politicians needed to identify the right metrics to ensure that the police were carrying out policy. I would argue that this is the wrong attitude. The problem faced by street-level bureaucrats is often that they are over-managed. Faced with the ‘problem’ of street-level bureaucracy, politicians and policy-makers have a tendency to try and tighten control. But, by the above account, street-level bureaucrats do not deliver on targets and policy because they are overwhelmed by the demands on their services. There’s only so much water you can wring from a rag before it runs out. Street-level bureaucracy is a phenomenon created by the relationship between policy-makers and those responsible for implementing policy: it is not solely the responsibility or failure of the workers themselves. Instead, policy-makers need to honestly engage with workers to understand their capabilities and needs.
I do not believe that this is an easy shift to make. Politicians are politically incentivised to promise to squeeze as much value as possible out of civil-service workers for as little money as possible. However, finding a way to do so would possibly have a positive impact on many policies.
2. Mitigate policy churn
It is literally the business of government to develop and improve public-policy. It is therefore difficult to be the person saying that you think policy should not be changed. Politics rewards people for shaking things up, not for having them remain the same. However, changing policies too rapidly can have a significantly negative impact on policy, a phenomenon known as ‘policy churn’. The problem with policy churn is that making real impact-full change takes time, and the time that it takes to make change is not particularly aligned with most democratic countries’ election cycles. Starting new programs is expensive, and if a project is not given sufficient time to have an impact then the investment can easily be wasted. Furthermore, it takes time for project workers to understand their responsibilities and duties, make connections with stakeholders, and all the other things that new projects demand. If continuity is not preserved, this valuable knowledge can be lost every time a program is rejigged or rebooted.
There are various ways that policy churn can be reduced. One example response could be to raise the threshold for eliminating old programs and starting new ones, such as requiring greater justification be given as to why new elements cannot be introduced into an old program.
3. Use both quantitative and qualitative assessment methods
Assessment in government is a tricky business. For better or for worse, most people who conduct assessments of government programs have a considerable stake in the conclusion of their assessment. For those working for the current party, be they politicians, civil servants, or assessment agencies, there is often a significant incentive for the program to be found successful. Political careers can live and die on the public’s perception of whether a project succeeded or failed, civil servants are much more likely to flourish if they enable the careers of politicians and their peers, and ‘independent’ assessment agencies may hurt their chances of being rewarded future assessment contracts if they are too negative in their reviews. Assessment in government, in short, is a perennial problem. My goal here is to address only a single piece of the issue, that being the importance of utilizing qualitative data alongside quantitative data when assessing projects.
If you’ve read my piece on policing statistics, you’ll know that I think the use of statistics can sometimes be problematic. Numerical data is not an objective representation of reality. Numbers are a product. They are created by chopping experiential data up and discarding most of the superfluous noise. The mechanisms we have created to so simplify reality are extremely valuable, but they can also bury theoretical assumptions, distract from the real-world objectives of policy, and encourage system gaming. Take crime statistics. While I was working in the Greater London Authority, there was a great fuss made about how arrest rates have been dropping in the city. The problem with this kind of knee-jerk reaction is that the fact that arrest rates are dropping tells us very little about anything. It could have been because of police budget cuts. It could also have been because the police were resolving more cases informally. The numbers, on their own, are insufficient to understand what’s actually going on.
My proposal is that projects and programs should utilize a combination of qualitative and quantitative data. By doing so, we both re-introduce consideration of the real meaning of the quantitative data, and remind ourselves that our objectives are the real, concrete states of affairs that numbers are meant to represent.
4. Remember external validity: focus on process rather than outcomes
The reasons that policy does or does not work can be difficult to identify, which makes it highly important that successful policy not be simply copied whole-sale. That a policy succeeded there does not mean it will succeed here. This is not to say that we should not learn from others, but rather that we must actively learn from others. We are much more likely to succeed ourselves if we attend not only to what others do, but also how they do it. What processes were used to survey the environment? Were other options considered? Why was the implemented solution chosen? This will not guarantee a policy’s success, but it does help mitigate the probability of policy completely missing its target.
Concluding thoughts:
One thing that all my principles have in common is that they often do not translate well into politics. I do not take these principles as easy or obvious, and I understand why they are often neglected in policy design. I do believe, however, that keeping these principles in mind would, in many cases, improve public policy.
Comments