Last Monday, the finance ministers of the EU-27 have discussed the necessary reforms to the Stability Pact. One of the sticking points at that meeting was apparently how financial penalties should be determined for the member states that would let them slip deficit and public debt because of lax economic policies. I quote a recent article Point published here so that everything is clear:
" Devoted to the issue of automatic sanctions and the criterion of reduction of debt, the meeting was to reconcile the positions of Germany, who advocates a hard line, that of the majority of states, including France, eager to preserve leeway.
Some question the overly automatic sanctions and the criteria to be used for examining the evolution of the debt. Others are reluctant to the very principle of financial penalties levied on EU funds "
In particular, Minister of Economy and Finance, Christine Lagarde said this:" France has always supported a strong and credible economic governance. From there to provide a fully automated nature, a power that would be totally in the hands of experts, not . In contrast, the German finance minister, Wolfgang Schäuble, as ECB President Jean Claude Trichet spoke in favor of a hardening material, in particular the German minister who publicly campaigned for sanctions "quasi-automatic ". Specifically, Mr. Schäuble has also insisted on the "automatic" penalties put in place for a debt or deficit exaggerated the lifting of sanctions to obey a "qualified majority reversed", that is to say that the penalties apply unless a majority of state are opposed. I confess that, initially, the position of this gentleman seemed much more sensible than the French position, being persuaded that the opportunity to vote on a discretionary basis for sanctions would limit the effectiveness or even completely negate mere possibility that they may exist.
So, to summarize, some member states, including France, defending the levels of penalties determined on a discretionary by a show-off states, while other states, not least because Germany is one of them, argue in favor of sanctions or quasi-automatic machines that would not rely on a political voting states. We're a little debate in the regulatory policies vs. discretionary policies ("rules Rather Than discretion, Kydland & Prescott 1977), the ultimate effectiveness of the system of sanctions is probably related to its credibility. However, it is not at all to this aspect I want to discuss today, but rather, more basic, the effectiveness of sanctions endogenous sanctions or exogenous (automatic) on fiscal restraint States.
This kind of situation in which states negotiate institutional arrangements that will govern them in the future lends itself to speculation or evidence. Especially here, where what is at stake is the effectiveness of specific devices of sanctions resulting from lax fiscal policy on the part of a Member State. In this case, it would be interesting for representatives of EU states who are currently discussing these points, have the chance to have a small presentation on the main results of the literature Experimental effectiveness of sanctions on the cooperation of agents. I will not do it now because it would take me too far, but just give some information on the following question: is it preferable to have the efficacy of automatic sanctions that would therefore not be chosen endogenously by Member States or, conversely, is it better that the award of sanctions against a Member State relies on a political process of voting discretion on the part of all Member States This case is based on having?
I will make a assumption that I understand it may seem questionable (and I have a problem too), but is particularly suited: the fact that each Member State shall respect the rules of the Stability Pact is akin to a public good that benefit all Member States, even though it may be individually expensive for a given Member State. Clearly, each state would benefit from being able to play stowaways and not follow the rules (make the budget deficit when he needs it, this significantly beyond 3%), but collectively, the effect would catastrophic, the absence widespread fiscal discipline leading to major economic problems for the eurozone. Just like the fact that a State which has committed at the international level does not negotiate with potential terrorists do it may quietly, prompting the terrorists to expand their business, to use the canonical example of Kydland and Prescott.
Tyran and Feld in 2006 led an experiment whose teachings are, if any, interest seems to me there. Their experiment, published in Scandinavian Journal of Economics , based a gambling problem contribution to a public good with penalties. The design of the experiment is however quite different from what is usually done on this issue. Indeed, many experimental studies on the effectiveness of sanctions in social dilemmas, and I helped myself, rely on endogenous penalties determined by individual participants who observe the level of contribution to public good of others. The canonical article on this kind of two-stage game, a contribution of sanctions and the following is the article by Fehr & Gaechter, 2000.
In their study, the situation is reversed. Participants first choose the terms of sanctions and then contribute. This experimental situation more or less corresponds to the current situation of the European Union, currently employed to decide the modalities of future penalties that would apply to Member States lacks fiscal discipline. It is therefore particularly interesting to note the results.
Penalties for insufficient contributions may be in some treatments, exogenous (that is to say, determined by the experimenter), and in others are endogenously chosen by a vote of the participants belonging to a group. In this case, the game is played in two stages: first the participants vote for the type of sanction they themselves attribute (no penalty, sanction or penalty moderately severe) and then, once the system of sanctions collective set, playing a game of classic contribution to public good (each member of a group of 3 has 20 tokens and must decide to allocate this endowment between a private account, and earns 1 point per token and a public account , and earns 0.5 points per token, but also reported 0.5 percentage point to the other two team members). Once the contributions chosen by each participant, the sanctions are applied. If the contribution to public good is equal to 20 (the whole allocation), there is no penalty. Otherwise, the penalty applies and takes two forms depending on the treatment, punishment and moderate to severe punishment. The penalty is to remove moderate gains 4 points in the participants did not contribute fully to the public good, and severe punishment is to deprive him fourteen points. For example, a participant who contributes 0 to the public good will in the case of treating "severe punishment", and assuming that the other two participants belonging to his group contribute fully, a gain of 20 +0.5 (40) -14 = 26 points.
From a theoretical perspective, the balance of such a game is the balance of free riding (all participants contribute zero to the public good) in case there is no sanction in the case where the penalty is moderate. In cases where the penalty is severe, the Nash equilibrium is an equilibrium has contrast in which the three participants contribute their entire endowment to the public good.
The main results are as follows. The chart below shows the average levels of contribution in percent (100% corresponding to a contribution of 20 tokens, which the Pareto optimal if everyone contributes here), this level of punishment for each:
source: Feld and Tyran, 2006
As Not surprisingly, the level of contribution increases with the level of sanction. But in fact, if sanctions are imposed exogenously (without a vote by the subjects), only severe punishment is efficient for the public good is produced at most. The penalty is moderate for its relatively inefficient, the level of contribution to public good is not significantly different level from a situation in which there is no penalty.
The results are quite different in the case of endogenous penalties, as shown in the following graph:
source: Feld and Tyran, 2006
The level of contribution is this time significantly even in the case of mild sanctions, equal to about 65% of the maximum possible, against only about 20% in the absence of sanctions. In the case of severe sanctions, it has no fundamental difference between contributions under this level of penalty to be voted by the participants or attached externally by the experimenter.
I will not say that every day, but it is clear, given these few results, I am compelled to agree with Christine Lagarde (arrghhh ...) It is better for sanctions policy as determined by the states Member of automatic sanctions, particularly if the level of sanctions is relatively moderate (one speaks of 0.2% of GDP). If sanctions are very strong, what seems to be the preferred option by Germany, then the procedure for determining sanctions is not very important. That it is a discretionary voting to decide to implement sanctions once observed behavior, or that one has an automatic sanction when a state member crosses the red line, sanctions will be effective to encourage States to contribute more to the public good.
However, I must emphasize that, as sanctions are costly, experimental literature emphasizes Detrimental effects of sanctions: the earnings of players can ultimately be lower if it is awarded on too strongly, and this may undermine the effectiveness economy. Besides, I already explained that such mechanisms can cause crowding intrinsic motivation of agents, extrinsic motivations (fear of the stick), replacing the intrinsic motivations (available natural agents to cooperate).
The sanction mechanism is handled with care and it is more than necessary to weigh the advantages and disadvantages.
PS: the picture that illustrates this post is taken from Gotlib, Topic in Brac, Volume 4 "Slowburn gag" with all my admiration and respect.
The level of contribution is this time significantly even in the case of mild sanctions, equal to about 65% of the maximum possible, against only about 20% in the absence of sanctions. In the case of severe sanctions, it has no fundamental difference between contributions under this level of penalty to be voted by the participants or attached externally by the experimenter.
I will not say that every day, but it is clear, given these few results, I am compelled to agree with Christine Lagarde (arrghhh ...) It is better for sanctions policy as determined by the states Member of automatic sanctions, particularly if the level of sanctions is relatively moderate (one speaks of 0.2% of GDP). If sanctions are very strong, what seems to be the preferred option by Germany, then the procedure for determining sanctions is not very important. That it is a discretionary voting to decide to implement sanctions once observed behavior, or that one has an automatic sanction when a state member crosses the red line, sanctions will be effective to encourage States to contribute more to the public good.
However, I must emphasize that, as sanctions are costly, experimental literature emphasizes Detrimental effects of sanctions: the earnings of players can ultimately be lower if it is awarded on too strongly, and this may undermine the effectiveness economy. Besides, I already explained that such mechanisms can cause crowding intrinsic motivation of agents, extrinsic motivations (fear of the stick), replacing the intrinsic motivations (available natural agents to cooperate).
The sanction mechanism is handled with care and it is more than necessary to weigh the advantages and disadvantages.
PS: the picture that illustrates this post is taken from Gotlib, Topic in Brac, Volume 4 "Slowburn gag" with all my admiration and respect.
0 comments:
Post a Comment