Public Policy Analysis: Implementation and Evaluation

We are going to ask ourselves the question of what happens once a law has been formulated and must be implemented and enforced, including by public administrations. We will proceed in three stages: firstly, by reminding us of what implementation is and how it is defined when we do public policy analysis, i.e. the general message is that it is a highly conflictual and open process that can potentially completely reverse what has been decided at the programming level; and then we will focus on a particular element by taking an example of the implementation processes in Switzerland that relate to what is called "polity", i.e. the federalist institutional architecture of Switzerland and the fact that the implementing federalism basically says that the Confederation formulates the laws, but it is the cantons and communes that apply. This particular configuration leads to difficulties and may also be advantages in terms of implementation. Thirdly, we will also show how important it is when analysing the implementation actions to really focus on the field, i. e. the behaviour of what are called field bureaucrats, street bureaucrats or bureaucrats at the counter. We will take entries by examples rather than by great theories.

Languages

English

Français

Implementation: an open and complex process

The studies on implementation date from the 1970s and 1980s in the United States. The seminal book of implementation studies is called Pressman and Wildavsky's Implementation, which explains the absolutely incredible mechanics of public policy and all the workings that are necessary for implementation to proceed smoothly. The subtitle is How Great Expectations in Washington Are Dashed in Oakland; Or, Why It's Amazing that Federal Programs Work at All, This Being a Saga... Morals on a Foundation. This subtitle asks about the gap that may exist between development and implementation on the ground. Various studies on implementation have been launched to take a fresh look at these implementation activities.

The almost canonical definition of implementation is a social process in which actors assert their interests, powers and possibilities of influence. Laws, ordinances and other prescriptions are to a large extent prescriptive offers that different adventures can use for varying purposes. It is a vision that stipulates that the content of public policy will depend on how the actors in the field who are responsible for implementation will draw on this normative offer and what they will do according to their own self-interest in what exists in the laws. If we analyze implementation, we really analyze the negotiated and strategic process of enforcement of existing laws, we talk about "implementation game" illustrating this description of implementation as a social, open and complex process.

Federalism of enforcement - diversity and deficits in the cantons

How does this translate into action in Switzerland? This is reflected in particular in the fact that cantons and municipalities generally have the power to implement federal laws. This is called enforcement federalism. The Confederation, the National Council, the Council of States and even the vote of direct democracies dominate in the formulation of public policies. Secondly, when it comes to applying, implementing and enforcing these laws, cantons are usually designated as responsible. This division of tasks, the federal government formulates and the cantons implement and often a little more complex than the clearly separated dichotomous approach presented. What is certain is that there is cantonal or even communal autonomy in the application of federal laws and this often leads to an extraordinary diversity in the implementation of the same law. So the same law applied in a canton or in another canton translates into implementing acts, therefore also into effects on citizens that are totally different, including effects in terms of the obvious unequal treatment of Swiss citizens.

We will see how classic analysis problems illustrate impacts on enforcement federalism. The question here is what is the impact of polity on public policy, i.e. what is the impact of federalism on the implementation of public policies?

Example - Wildlife regulation / hunting (cf. Nahrath study, 2000)

The first example is a current example of how hunting is regulated in Switzerland. If we look at how hunting and wildlife are regulated, there is a fairly impressive diversity and what we call a "federalist laboratory", which means that the cantons are experimenting with different ways of regulating hunting. Why? The federal government says in the Constitution that there is a public monopoly on the allocation of hunting rights, but it is up to the cantons to determine which hunting regime they will actually be able to implement in order to regulate who can hunt, but also with what intensity. The Confederation limits itself to certain things such as protected species, (we do not have the right to shoot anything), usable weapons, or the use of traps; when we analyse public policies, we enter very different areas. So how did the cantons decide to regulate hunting, in view of the great room for manoeuvre afforded them by the Federal Constitution, and therefore by federal policy?

This map of Switzerland shows that there are three main ways of regulating hunting. There is a national exception in Geneva where hunting is prohibited. There is a system of more alpine Romanesque cantons and the two cantons of Appenzell Inner and External Rhodes in which the regulation mode is the permits, so to hunt it is necessary to have a permit, and finally the Swiss German-speaking cantons have opted for a third type of management mode which is called "leasing".

In the case of the canton of Geneva, there is a ban on hunting for private individuals. Only gamekeepers who, since the adoption of this regulation in Geneva, can actually kill animals. This translates into very high costs, since game wardens must be hired to regulate wildlife. This is also reflected in the absence of perception of the hunting regal in the sense that no licences are sold, so there is no financial income; and this is also reflected in the effects that some have denounced as a poorly conducted hunting ethic. Hunting ethics brings together the main principles that hunters should respect, such as, for example, the fact that it is not necessary to shoot at night with infrared rifles. This is prohibited by law even at the federal level, but in Geneva, if we hunt during the day, there is a risk of accidents, so they shoot at night without complying with the law, but that is the price to pay to actually implement this ban on hunting.

Under permit systems, there is regulation through licensing of different hunters and these hunters will go to the Alpine cantons or take a permit to hunt. So access to hunting will be regulated through the granting of licences. These licences are generally not regionalised in the sense of obtaining a licence and can be hunted freely throughout the cantonal territory. Certainly, there are sometimes maximum quotas that cannot be exceeded, so we cannot kill more animals than the quota provides. In a majority of situations, the problem right now is not that we kill too many animals, but that we don't kill enough animals. There is an under-exploitation of wildlife and wildlife causes problems, particularly forests and crops such as the case of wild boar in the canton of Geneva. This system of management by permit does not really allow, contrary to the Geneva system for example, a balanced management of wildlife throughout the territory.

The third model that we find is the leasing model, which is totally different from the idea that we don't allocate individual permits, but we attribute the management of a territory for 6 to 8 years to a hunting association and it is it that during this period will regulate the number of animals that we can draw and it will also be responsible for the damage done by the fauna to the forest or to the crops and crops. If a hunting company has mismanaged the given territory for 6 to 8 years, then we may lose our lease and be excluded from the practice of hunting.

The empirical analysis of these three systems has shown that in terms of sustainability of rational management of the "wildlife" resource, it is generally the leasing system that proves to be the most efficient. It is by empowering an association of hunters on a territory rather than giving bureaucrats the possibility of killing or individuals, through a licence, the possibility of killing animals that we have the best management of the "wildlife" resource.

These are therefore three systems based on the same delegation of powers at federal level, which are really very different in terms of the way they are implemented, but also in terms of their effects. That's what execution federalism typically allows. It makes it possible to experiment in a federalist laboratory with different application solutions and, in the case of hunting, this has not really proved to be very promising in the sense that everyone would have learnt about what the neighbours are doing and the more or less positive effects of their way of regulating, but in some other public policies, when there are such marked cantonal differences, we experiment with ways of managing, for example, drugs and tobacco. In drug policy or in various social policies, implementation federalism has allowed diversity, but also experimentation and ultimately the choice of the best solution.

Example - Acquisition of real estate by foreigners (see Delley et al., 1982)

This second case relates to the classical analysis as done by Jean-Daniel Delley (law professor, UNIGE) on the acquisition of real estate by foreigners. This law that is still in place and this policy that is still in place began in the 1960s when we realized that more and more land, but especially buildings, had been acquired or wanted to be acquired by foreigners.

Some movements at the time said that there was foreign control over Swiss territory and that this represented a real danger to the national interest. It was already at the time a rather nationalist right-wing that wanted to fight against the growing foreign influence on Swiss soil.

It was not the only ones to focus on this issue, there was also the left which was concerned about the phenomenon observed, but for a very different reason, a completely different framework which was for social reasons, namely the fact that there was foreign capital which bought buildings to speculate on real estate and which therefore contributed to the scarcity of certain moderate rents or even to the increase of all rents in Switzerland. Finally, there was a third actor at the time who was participating in the debate, which was the rest of the right-wing liberal right-wing, who said that there is absolutely no problem or regulation at all, since it is necessary to defend private property, the free trade tradition of Switzerland and if a private owner is foreign, that does not pose any problems in itself. Therefore, there is no state regulation of the creation of certain buildings or the control of rents, nor is there any danger of foreign control over the land.

These three forces were fairly balanced, there was no clear majority for one of the three parties, so a compromise was reached that allowed for all the policies. It has been said that this problem must be effectively controlled, that foreigners wishing to acquire real estate in Switzerland must have a State authorisation, that it must be formally authorised by the State, but the question is under what conditions can it be granted. In the law, authorization is granted if there is a legitimate interest. It is an indeterminate legal concept. The outgoing legislation had to be implemented by the cantons. Therefore, it was the cantons that had to grant/refuse permits to foreigners who wished to acquire a building that had the responsibility of defining what a "legitimate interest"meant in practice. The law pursued very different objectives, there was not a solid compromise at the level of the Confederation, so, inevitably, at the level of implementation, there was an instrumentalization of the federal law according to purely cantonal and local interests. Some have talked about the cantons hijacking the law.

This graph shows that there is an initial situation with cantonal practices which are to some extent fairly similar or comparable, and then these various laws are introduced, which try to reduce or at least control the acquisition of real estate by foreigners. The different cantons have chosen completely different strategies.

The canton of Valais has said that it was a policy of tourism support thanks to foreign capital, so the stations will be cemented with the influx of foreign capital, so Valais has a policy of tourism expansion and anyone could obtain an authorization even when the Confederation questioned certain authorizations and legitimate interest. The parade was to create local trustees with foreign capital to effectively enable a pseudo-Swiss to acquire a property on behalf of a foreigner.

On the contrary, a tourist canton such as Lucerne, in which the local political forces have used this law as an anti-foreign law that does not want to lose control over tourism development, must remain in Swiss capital. If we look at the authorisation curve in Lucerne, it is about zero. They really had the opposite enforcement.

In the case of Geneva, there was a policy that basically fluctuated with developments in the housing market.

So, this example is really a classic example, which is one of the first studies on the implementation of public policies that were done by lawyers who wondered what happens to the laws that we adopt, especially when they are formulated in a very vague way, in a very broad way, for example through the concept of legitimate interest.

Example - Disability Insurance Reforms (see Byland et al. 2015 study)

If the law is formulated in a much more concrete and precise way, perhaps there will be fewer variations and fewer differences in implementation between the cantons.

Disability insurance reforms: unequal treatment or convergence between cantons?

Disability insurance reforms: unequal treatment or convergence between cantons? aThis curve represents the demand for disability insurance between 1990 and 2011. In bold, the disability insurance deficits that followed a roughly inverse curve are shown in bold, except that there is an inflection point. Obviously, there was a policy that was introduced to try to reduce the granting of annuities and thus reduce disability insurance deficits.

If we look at how this disability policy has evolved, there are three main periods: 1999 - 2003,2004 - 2007 and 2008 - 2011. Each of these periods is characterized by a particular public policy. In the first case, the instrument of public policy was the granting of pensions, that is, someone has a disability, someone who has a disability can no longer work and they are paid a pension. This was the classic policy, the one that encouraged many people to apply for pensions and which was reflected in the growth of the deficit, or at least the increase in disability insurance deficits. Faced with this observation, it was decided to introduce a reform in 2003 in which it was decided that pensions would be granted, but that if we could not get people back into the labour market. So we always give pensions, it's the main instrument, but we still try to keep people with disabilities in the labour market. For example, they were granted partial pensions, i. e. they received a pension, but they still work part-time to keep them in the labour market and this reduces the pressure on disability insurance. This measure proved to be insufficient to clean up the disability insurance budgets, which meant that there was still a further restriction on disability insurance, where it was said that we were going to change our philosophy and the annuity is only a last resort. It is necessary to prove that there are no opportunities in the labour market, that not all measures have been taken to re-enter the labour market before receiving a pension. This is actually a restriction in the disability insurance policy. The longer we move forward, the more difficult it becomes to get a pension because the longer we wait for us to work and still be included even part-time in the labour market.

How does this federal policy, which is applied at the cantonal level, translate? The cantons have AI offices. An application is made and the AI office will accept or refuse, partially or totally accept the pension application. How this policy is implemented in the various Swiss cantons.

Rate of allocation of DI pensions (IV offices) by canton and plan.

The three periods are shown here and the national average of the annuity acceptance rate is shown. The longer we advance in time, the more we see that the rate of acceptance of pension applications falls, so there are fewer and fewer pensions that are accepted compared to the applications that are formulated. The political scientist's question is whether you have the same chances of getting an annuity if you are in one canton or another, or whether implementation through the cantons will discriminate. The same situation can lead to totally different decisions by the cantonal administration under the same federal law.

We often talk about Lake Geneva syndrome when we say that only the Romanesque cantons pay pensions. The most restrictive canton is Valais, the canton of Neuchâtel has experienced a restriction in its practices over time. What we observe is that all cantons are becoming increasingly restrictive, but at very different rates. As a result of these successive reforms, there seems to be more diversity in the implementation by the cantons and possibly more unequal treatment between the different citizens who are in different cantons.

Rate of AI pension provision by AI boards per plan (median and quartiles).

The box-plots with the median, first quartile, second, third and fourth quartile are shown here. The dispersion observed at the beginning is initially reduced and increases significantly in the third period. Today, political inequalities in the political sense of the term appear to have increased between cantons. This is a purely quantitative analysis, it would of course be necessary to compare comparable files in different cantonal systems in order to see whether, in fact, it is objectively similar situations of the disabled that lead to totally different decisions.

Example - Snow Cannons

This example highlights the differences between what is provided for by law and what is actually implemented, even where the rule is very clearly defined. This is referred to as implementation deficits and misappropriation of the law. In other words, there is a law that exists, it is very clear, but it is circumvented at the local level. Sometimes, at the local level, implementation runs up against configurations of actors that simply make it impossible to apply the rules. The balance of power is such that there is a proper circumvention of the law.

In this case, we really have artificial snow cover that is regulated and very clearly defined. There is a master plan that sets a standard that does not suffer from any interpretation, which is the fact that there is no right to snow before November 1. This law arrives in a field in the Valais in the commune of Agettes which is a small town that should merge with Sion, ecological environments that try to enforce artificial snowmaking only after November 1st, a local promoter of Téléveysonnaz and then actors who are the hydroelectricity producers who need water to turbine and produce electricity and who would not want their electricity to be supplied by electricity.

The power relations are so strong in favour of the target group of this public policy, i. e. Mr. Fournier being described as an influential tourist actor and someone above the law and it can, in all serenity, by announcing it without too much eyebrows, as a given fact, but that there is a law and that we bypass it because it is quite natural to bypass it. What is most surprising is the passivity of the municipality and also the passivity of the ecological circles which obviously did not prosecute this illegal act before the courts.

As far as implementation is concerned, there is no request for authorisation, there are no reactions from the municipality that should have granted it or denunciation to the cantons, and there is tacit agreement between all the players to say that the law is certainly not being applied.

This example is symbolic and speaking. There are cases where it is a little more awkward because the stakes are higher than snow than a few ski runs. We must retain the concept of implementation deficit, implementation gap, implementation gap when, truly while the law is clear, resistance at the local level prevents any activation of legal norms.

Bureaucratie « de terrain »

Article détaillé : La théorie de la Street-Level Bureaucracy (SLB).

Définition

Il n’y a pas uniquement le fédéralisme d’exécution qui est un facteur explicatif des difficultés que l’on retrouve au niveau de la mise en œuvre, il y a aussi le fait qu’au sein même de l’administration, il y a certains acteurs que l’on appelle des "street-level bureaucrats", des agents de première ligne dit aussi des agents au guichet qui jouent un rôle décisif dans la mise en œuvre. Ces bureaucrates dits "de terrain" sont généralement des membres de l’administration qui sont en interaction directe et régulière avec les citoyens auxquels ils donnent des autorisations ou ils fournissent des prestations. Mais, dans l’exercice de leurs fonctions, ils bénéficient, d’une chose essentielle qui est le pouvoir discrétionnaire. Les règles qu’ils doivent appliquer ne sont pas suffisamment précises afin de leur dire comment se comporter dans tous les types de situations. C’est donc eux qui en fonction de leur marge d’appréciation vont décider de la manière dont sera mise en œuvre la politique. De plus, ce pouvoir discrétionnaire peut s’exercer parce que bien souvent, ils bénéficient d’une certaine autonomie vis-à-vis de leur hiérarchie.

Si on applique ces caractéristiques, à savoir interaction directe régulière, pouvoir d’interprétation et marge discrétionnaire, cela s’applique à de nombreux fonctionnaires dont certains peuvent jouer un rôle important. Ce sont par exemple les enseignants ou professeurs d’université qui sont des bureaucrates de terrain, mais aussi les policiers, les travailleurs sociaux, les juges, tout le personnel de la santé publique qui fournit des prestations, bien souvent, le nombre d’actes qu’ils doivent poser pour mettre en œuvre une politique publique sont tellement diversifiés que la loi ne peut pas prévoir toutes les situations concrètes, individuelles auxquelles ils seront confrontés. C’est donc à eux de juger sur le moment que faire, prendre une sanction, ne pas prendre une sanction, délivrer un soin, ne pas délivrer un soin, choisir un thème ou ne pas choisir un thème, etc. Lorsqu’on formule une politique publique, on est tout à fait conscient que l’on va pouvoir buter lors de la mise en œuvre sur la résistance des bureaucrates de terrain.

La thèse de Lisky qui est la personne à l’origine du terme « street-level bureaucracy », dans son ouvrage Street-Level Bureaucracy: Dilemmas of the Individual in Public Service publié en 1980 est que « I argue that the decisions of street-level bureaucrats, the routines they establish, and the devices they invent to cope with uncertainties and work pressures, effectively become the public policies they carry out. I argue that public policy is not best understood as made in legislatures or top-floor suites of high-ranking administrators, because in important ways it is actually made in the crowded offices and daily encounters of street-level workers ». Il faut se concentrer sur les interactions de ces bureaucrates de terrain avec les citoyens parce que c’est là que se joue et se noue la politique publique. En fait, la politique publique est vraiment les décisions et les actions que prennent ces bureaucrates de terrain. C’est la manière dont ils développent des routines ou des stratégies afin de mettre en œuvre les politiques publiques dont ils sont responsables.

Exemple d'une réforme de politique sociale en Californie (cf. étude Meyer et al., 1998)

Cet exemple est celui des personnes qui étaient à l’aide sociale aux États-Unis. La réforme avait comme destinataires les personnes qui étaient récipiendaires de l’aide publique de l’assistance sociale et qui touchaient généralement une rente ou une aide financière dès qu’ils remplissaient certaines conditions comme s’ils n’avaient pas de travail, étaient dans une situation familiale difficile ou dans un quartier à problèmes. Le but de la réforme que l’on a voulu introduire en Californie qui s’inscrit parfaitement dans les buts d’activation dans le marché du travail dont on parle depuis une dizaine d’années dans tous les pays européens aussi était d’essayer de les réintégrer, de les réinsérer dans le monde du travail donc de promouvoir ce que l’on a appelé le workfare plutôt que le welfare. On promeut l’État social actif ; on est prêt à apporte une aide financière pour autant que le récipiendaire fasse aussi un effort pour se réintégrer et pour prendre toutes les mesures afin de réintégrer le marché du travail en particulier.

Lorsqu’on a souhaité introduire cette réforme, on s’est dit qui va concrètement la mettre en œuvre. C’était les bureaucrates de rue qui étaient des gens qui travaillaient dans des agences sociales à un niveau décentralisé dans des contés en Californie et on été tellement conscient que toute la réforme n’allait avoir d’effets que si ces gestionnaires dans les services sociaux jouaient le jeu qu’a été mis en place tout un programme de formation afin de leur expliquer qu’une nouvelle réforme était souhaitée, quel était le sens de cette réforme et surtout comment devaient-ils se comporter par rapport aux gens qui étaient récipiendaires de l’aide sociale. Donc, de manière ciblée on a vraiment insisté sur la formation, la sensibilisation, la motivation de ces bureaucrates de rue.

À été essayé, après avoir mis en œuvre la réforme, de regarder comment cela s’était passé. Les personnes qui ont analysé la mise en œuvre ont observé au travers d’approches ethnographiques comment les travailleurs sociaux discutaient avec les récipiendaires de l’aide sociale et est-ce qu’ils ont essayé de les inciter à se réintégrer sur le marché du travail. Ils ont regardé dans quatre comtés représentatifs au travers d’une observation directe des interactions de face-à-face entre la personne qui travaillait pour l’État et celui qui venait demander de l’aide sociale et ils ont également fait les entretiens après avec ces bureaucrates de terrain pour essayer de comprendre ce qui se passait. La surprise ou plutôt la désillusion était que lorsqu’ils ont fait ces observations, jamais, le mot d’ordre qui avait été donné, "it always pays to work", n’avait été mentionné par les bureaucrates de rue. Donc, la réforme était réduite à néant ; malgré la formation, malgré la sensibilisation, rien ne s’était passé.

À l’aune de ce cas, il faut retenir que vouloir mettre en œuvre une politique publique présuppose la collaboration, y compris des plus petits fonctionnaires que l’on a dans la hiérarchie. La politique d’éducation présuppose que les plus petits fonctionnaires de l’éducation comme, par exemple, les professeurs jouent un rôle pour appliquer une réforme. S’ils résistent, s’ils s’opposent à cela, ces petits bureaucrates de rue peuvent bloquer tout ce qui a été décidé en amont et y compris au niveau du parlement.