Is Switzerland Losing its Precious Heritage of Decentralization in the Name of Efficiency?

Victoria Curzon-Price

Switzerland has one of the highest per capita incomes in the world in comparative terms adjusted for local purchasing power. It is currently in fourth place, after Luxembourg, USA, and Ireland. However if you look at the real growth rate, Switzerland ranks 21st out of 22 of the OECD's most advanced nations, well down on the list. Something is wrong.

People are saying: “We don't feel so rich. In fact, it's harder and harder to pay the bills at the end of the month.” Why should this be so? People offer many explanations: too many taxes, too many regulations, building permits too difficult to get, too difficult to start a business, too much immigration, too much emmigration — too much of this, too little of that. Everybody has their favourite reason — and their favourite solution. It is very difficult to choose between them.

Furthermore, Switzerland is one of the most democratic countries in the world and public decisions are very slow. The question therefore is: if it takes such a long time to agree on a solution to a problem, because of Switzerland's decentralized democratic machinery, does this mean that Switzerland has to sacrifice direct democracy, the referendum, public initiatives, rights of appeal against public decisions, in short, Swiss federalism — in order to become more efficient? In other words, has Swiss democracy become too expensive to run? Is part of the reason for poor Swiss growth not to be found in its decentralized political system?

My answer is “No”. The efficiency argument is used constantly in order to promote centralization and simplification of the political process. I would like to convince you that Switzerland has been a victim of creeping centralization. Centralization is part of the problem.

Thus, if you are looking for solutions to a real public policy problem, how do you choose between the hundreds of possible alternatives? The only way to discover the least bad solution to a public policy problem is by experimentation. But experiments in laws and institutions take a long time if all you can rely on is your own experience. You have to identify the problem, find a solution, test it in practice, see if it works, and correct it if it doesn't. The probability of getting in right first time is virtually nil. So the process must repeated again, and again, and again. This is the discovery process by trial and error.

For example, everybody in France agrees that the 35 hour week is a disaster: it was a bad solution to a correctly identified problem (unemployment). Everybody agrees that it doesn't work, but nobody dares do anything about it. Well, that is an example of how difficult it is to change things politically, even when you know you have made a mistake. Bad policies are much more expensive than slow, democratic, decentralized decision-taking procedures!

So if the only way to discover the least bad policy is by experimentation. Switzerland has inherited a treasure from it's history in the shape of its 26 Cantons, and its 3000 Communes. Here is a system of competitive experimentation in laws and institutions, here are 26 laboratories of public policy experimentation, and in some areas, 3000 even smaller ones. They are all facing similar problems, and are searching for solutions on a broad front. The probability of one of them coming up with a least bad solution is quite high. Then the others look and say: “Ah, that's a good idea. It seems to work. Let's try it too.” So the least bad solution spreads — until it is replaced by a better one! Because this dynamic process in fact never rests. There is no “optimum” solution, valid for all time. There is only an optimum system.

It happens half consciously and half rationally. It is not a pure market process of competition. It is midway between rationalism and the spontaneous market process. It is an efficient way of developing new laws and institutions to fit new needs as they appear. I think it served Switzerland well in past, which is why the current trend toward centralization is so worrying. Let me give you a few examples of growing centralization in Switzerland.

  • The share of the federal government in total tax revenue has risen in the past 25 years from 28% of the total tax take to 32%. You might say it is not a big change, but it is the direction which is significant. While Switzerland was gradually increasing the role of the centre, other countries in Europe were going in the other direction and devolving more power to regions. As a result, Switzerland is no longer so different from other countries in this regard.
  • Social security has risen as a share of total compulsory state revenue from 29% to 34%. As most social security is run from Bern, this suggests greater centralization.
  • Federal debt has risen from 10% of Switzerland's gross domestic product in 1970 to 28% today; Cantonal debts have risen from 10% to 16% of GDP during the same period. Total indebtedness does not look too bad from the outside — it looks much better than France or Germancy (each over 60% of GDP), for example. But that's not the point. It is the centralizing trend which matters and the trend has been alarmingly fast. Federal debt has risen in absolute terms from 10bn Swiss francs to 120bn Swiss Francs — a sign that federal expenditure is out of control.
  • Because in Switzerland, as in many other highly developed countries, politicians find it much easier to raise taxes than to cut expenditure, we are sure to be voting for higher federal taxes very shortly.
  • Finally, I would like to take as an example of creeping centralization the fiscal harmonization the law of 1990 (Loi fédérale sur l'harmonisation des impôts directs des cantons et des communes du 14 décembre 1990): According to this legislation, the 26 cantonal laws defining the fiscal base were to be harmonized throughout Switzerland, so all cantons (and the federal government as well) would use the same definitions for raising revenue. Cantons would remain free to set their own tax rates, but the tax system would be harmonized. The argument was that it would make the Swiss tax system more transparent, hence more efficient. This seems so obvious that it needs no further explanation. For example, I was struck by a public opinion poll organized by the federal fiscal administration in Bern. They asked people whether they would be in favour of harmonizing tax rates across all Cantons and 60% replied that it would be a good idea because it would be “more efficient”.(2) However, this is simply not true. It assumes that when the harmonization was decided upon in 1990, our political leaders had hit on the best possible solution for all time. But as we have just seen, the probability of finding the least bad solution first time around is very low, and the probability that the bad solution actually found will become more appro- priate over time is even lower.

Let me point out why, particularly in Switzerland, this centralizing tendency is most dangerous. The discussions on harmonizing the tax base started in 1968 and the law was actually passed 22 years later, in 1990. It was put into operation in 1992 and the Cantons have been given eight years to adapt their laws. Some are taking longer, so we are even now still in the process of harmonization. The whole process has taken 35 years. But do you think that law is perfection? Can it not be improved upon? Is it really adapted to the 21st century where we are challenged by globalization and Estonia has a zero tax on corporate profits? I would suggest there might be room for improvement. But if so, how do we agree on what improvements should be made? Will any change also take 35 years?

So you see, centralization is very dangerous for Switzerland because once the things are centralized it takes a long time for them to change and the world doesn't wait for Switzerland. What if this law harmonizing the fiscal base had not been adopted? There would have been 26 laboratories experimenting with different types of taxes each one searching through the forest of possibilities. The Swiss tax system would have been changing at the margin all the time. It doesn't matter if nobody finds the best solution, because today's best solution is tomorrow's burden. We would have had a flexible system for creating fiscal law. Instead we have got a system which can't be changed without huge political cost. For Switzerland, centralization is much more dangerous than for most countries, because of its excellent, highly decentralized, political institutions.

So which do we prefer: a centralized, static system which will yield growing inefficiencies over time, or a decentralized, dynamic system which generates efficient solutions all the time?

In my humble opinion, the solution to the Swiss problem of lack of growth is in fostering and preserving those 26 laboratories of experimentation in laws, regulations and taxation, because today a country's comparative advantage lies in this domain. Switzerland still has this heritage, but is fast losing it, too often fooled by the static efficiency argument in favour of centralization. This must be energetically opposed and replaced by the the vision of a flexible and dynamic system capable of constant change and evolution.


Remarks:

(1) Readers will recall Hayek's distinction between phenomena of human origin that lie between instinct and reason, like language, tradition or rules. The spontaneous order of the market also lies between instinct and reason. Here I am recalling that, as Hayek himself said, in questions of institutional evolution, we combine reason with spontaneous processes. In other words, we consciously imitate one another's institutions, thus accelerating human institutional evolution.

(2) Thierstein, Alain et al. “Les différences spatiales des charges fiscales et compétitivité régionale”, Société suisse d'études pour l'organisation de l'espace et la politique régionale OEPR, Zürich, 2003.


Victoria Curzon-Price is a Professor of Economics at the University of Geneva and President of the Mont Pčlerin Society.

2005

politische-vielfalt category logo