Theorie Part 3

• What to do if these aims are in conflict?

• What instrument(s) should be used to reach which target?

• „Golden Rule“: Policymakers must have at least as many independent policy instruments as there are independent policy targets.

• Tinbergen‘s so-called Golden Rule of economic policy is still valid as long as economic system is parametric.

• Assume a policymakers that has a quadratic objective function and faces linear policy constraints.

• Deviations from target values lead to losses (loss-function). Quadratic objective means increasing losses in deviations.

• Most of the times, policymakers will have more targets than they have independent instruments.

  • Central banks usually aim at low inflation and stable / high employment. Only instrument is monetary policy.

  • „Magic square“ has four targets which are mostly discussed with the optimal use of fiscal and monetary policy

• Introducing more instruments is often not possible and dropping targets may be politically unattractive.

• Even if policymakers have fewer targets than instruments, a problem may arrive if their use is not coordinated.

• Assume for instance that both government and central bank aim to keep inflation low. Central bank uses monetary policy, government sets fiscal policy.

• Even if both have similar objective function, uncoordinated policies can create problems.

• Solution:

  • Drop one instrument.

  • Coordinate policies.

• A problem arises if the instrument are not linearly indepent, so the instruments affect the two targets at the same time.

  • Monetary policy does not only affect output but also inflation, or the current account.

  • Fiscal policy increases output but also pushes up deficit.

  • Minimum wage may increase income equality but also increase unemployment.

• Golden Rule implies that number of targets and number of instruments should be equal.

• What are possible relations between targets and instruments?

• How to choose between different instruments to use a given target?

A target refers to an assigned value of an economic variable that is intended to be achieved. This can include prices, price indices, rates of change, quantities, or any other relevant economic measure. Targets can be set based on their intrinsic value or because they impact other variables or principles such as distribution and fairness.

Instruments, on the other hand, are variables that policymakers can directly or indirectly influence to achieve the desired targets, such as investments in education or research and development subsidies.

=> in most cases number of targets > number of instruments (Golden Rule likely to be violated)

=> the higher the number of targets, the higher the probability that they are inconsistent

=> instruments often do not influence just one target, they have often influence on other targets, too.

• Even if number of independent policy targets and independent policy instrument is equal, the question is which instrument to use for which target.

  • Basic idea is that instruments should be selected by their comparative advantage. They should be applied to that target where their marginal influence is largest. This is especially important if the use of the instrument is not costless.

  • Optimal assignment of instruments to targets implies that the marginal impact of all instruments should be equalized.

• In most cases, policymakers cannot be certain about state of economy, there might be measurement problems (or lags), demand or supply shocks, or uncertainty about the quantitative effects of a policy measure.

• Depending on the type of uncertainty, its existence will have a more or less significant effect on optimal policy.

  • We distinguish between additive and multiplicative uncertainty.

    Additive uncertainty does not affect optimal policy („certainty equivalence“).

  • Multiplicative uncertainty („parameter uncertainty“) does have an effect.

• The Tinbergen-Theil approach to economic policy assumes that policymakers face only technical problems in reaching their policy targets. If they have enough instruments, they can always reach their targets.

• Even if they have not, an optimizing politician can reach second best by trading-off his targets and accounting for uncertainty.

• What this neglects, however, is the interaction between policymakers and private agents who also influence the economy through their actions.

• If government and private agents have different objectives or information, a strategic interaction is created between them.

Lucas' critique questioned the assumption made in traditional econometric models that the economy's parameters remain fixed when policy changes are implemented. In reality, if individuals optimize their behavior, these parameters are likely to change in response to expected policy changes. This means that economic policy cannot be viewed simply as a matter of estimating parameters and predicting outcomes, but rather as a complex process involving strategic interactions between policymakers and the private sector.

=> Economic policy is no longer made in a parametric context but subject to strategic interaction between policymakers and private sector.

=> this interaction between government and private sector can be understood as a “policy game” in a game theoretic set-up

• “Rational expectations" are model-consistent expectations. Agents "know the model" which ensures internal consistency in models. To obtain consistency within a model, the predictions of future values of relevant variables are assumed to be the same as those of decisionmakers, given their information set, the nature of the random processes involved, and the model structure.

• Rational expectations imply that agents' expectations are correct on average and not systematically biased.

• Note that rational expectations is an assumption of aggregate consistency in dynamic models, it does not necessarily imply rational choice of individuals

• The private sector, with rational expectations, recognizes that optimal policy may change over time and anticipates actual policy based on available information.

• This can result in a suboptimal outcome for both the policymaker and the private sector, although it may be the only time-consistent equilibrium.

• As a result, it is argued that policy should adhere to a predetermined rule rather than allowing for discretionary decision-making.

• Under specific conditions, limited discretion may still be viable.

• Consequently, the debate revolves around the advantages and disadvantages of "rules vs. discretion" in policymaking.

• Assume that only labor is used for production.

• Production and output depend on the input of labor. Firms maximize profits and condition labor input on wages, which they take as given.

• Labor sets its nominal wage demand depending on real wage target and expected price developments.

• Monetary authority has preferences over output (employment) and inflation. Its only instrument is monetary policy, the use of which is costly since there is conflict between low inflation and high output.

• There is a strategic interaction between firms and wage setters, and a strategic interaction between monetary authority and private sector. We assume Nash-behavior.

• Strategic interaction exists between the private sector (wage setters) and the government (monetary authority).

• The private sector seeks to avoid surprises that reduce real income, while the government may use surprise inflation to increase output/employment.

• Individual expectations (𝜋𝑡𝑒) and mathematically correct expectations (𝐸𝑡−1[𝜋𝑡]) may differ unless there are surprises.

• The government faces a trade-off between stabilizing output and preventing inflation.

• Using inflation as a policy instrument to achieve output targets incurs costs.

• Based on this insight, government could announce that its policy targets for inflation and output would not be pursued since they are only creating losses. If credible, this would lower expected inflation to zero as well.

Summarized:

This is about the interaction between the private sector (workers) and the government's monetary authority. They have conflicting goals: workers don't want surprises in inflation as it affects their real income, while the government may use surprise inflation to boost output and employment.

We distinguish between individual expectations (what people think will happen) and mathematically correct expectations based on available information. Rational expectations assume they are equal, unless there are surprises.

The government faces a trade-off between stabilizing output around the target and avoiding inflation. Using inflation as a policy tool to reach economic targets comes with costs.

• A credible rule based policy can solve time-consistency problem and reduce inflation bias (possibly to zero).

• A credible rule is one that takes away monetary policy discretion completely. (Friedman rule, gold standard, currency board...)

• However, it comes at the cost of losing the possibility to stabilize shocks.

  • Under what circumstances are rules preferred?

  • Is there a way to improve upon strict rules?

  • Can credibility and restricted discretion be combined?

• Assume government has the possibility to cede the use of monetary policy. This could be a fixed exchange rate to another currency with non-active policy, or it could be the adoption of a policy rule with requires zero inflation.

• If commitment to the rule is credible, expected inflation adjusts to this lower level (here zero) and the inflation bias disappears.

• This comes at the costs of not being able to stabilize economic shocks anymore. This trade-off has to be considered when rules dominate discretion.

• Since decision is taken at begin of period, government minimizes expected losses before potential shocks are realized.

• We look at one period only to simplify analysis.

• Objective function:

• Credible commitments are very costly because they imply giving up policy options in each and every case.

• Besides being very expensive, they might not be fully credible because government can almost always break its commitment (usually at the cost of losing reputation or being voted out of office).

• Costs of breaking a commitment should make it less likely that this occurs, but at the same time it is still a possibility to do so, if the alternative is „bad enough“.

• A conditional commitment can be seen as a contingent rule. In normal times, the commitment is honored. In real bad situations, however, one can revert to discretion and „escape“ from the policy rule.

• The effect of such an escape clause critically depends on the timing of interaction between government and private sector.

• if expected inflation is a function of the critical shock, and the critical shock is a function of expected inflation, then there is circularity between the two variables that does not allow for a unique solution.

• Thus, arbitrary changes in expectations can bring about a policy change. There are multiple equilibria with self-fulfilling expectations

• Objective Function:

• If private agents know that governments can break their commitment to a rule, governments have an interest in establishing a reputation for not doing so.

• In fact, many agents form expectations about actual policy on the basis of observed policy. If government „behaves well“ they expect it to continue to do so, but if it once reneges on its promises expectations will adjust and „bad behavior“ will be expected for the future.

• Government aims to establish reputation for „behaving good“.

• This is related to, but not equal to, the considerations before. Here we assume that private sector is willing to believe government (and adjust expected inflation to zero) but punishes government if expectations are not fulfilled.

• Expectations and behavior are a combination between adaptive and rational.

• Explicit consideration of multiple periods (dynamic game) and the importance of time horizon.

• We use simpler model than before. Utility is a combination of linear and quadratic objectives (linear-quadratic model), and we abstract from shocks.

• Objective Function:

=> „Reputation“ works very badly in theory, but may work in practice

• The solution mechanisms to the time-consistency problem considered so far have tried to change government behavior by punishing government.

• Another possibility would be for government to delegate policy to an independent agent whose incentives are different. Different preferences will in general lead to different policies, and by choosing the „right“ agent or setting the „right incentives“ the policy outcome improves.

• Conditional on private sector knowing about different preferences/incentives and delegation being credible and nonrevocable

• Ideally, delegation alleviates the time-consistency problem and makes it possible to achieve a better trade-off between rules vs. discretion.

• Giving that the agent enjoys a higher credibility, it should be possible to combine (on average) less active policy with some flexibility.

• Strict rules are substituted with „flexible“ rules.

  • Delegation to international organization, like EU, IMF, World Bank…

  • Delegation to independent agency, such as central bank, judicial courts, competition authorities…

• A possibility of restricted delegation is the case where an agent is ordered to implement a policy whose target is specified by the government.

• Agent is free to use instruments as he prefers but must report and justify failure of reaching the policy target.

  • Gives principal some control of agent‘s behavior.

  • Still makes use of agents expertise.

• Example: Inflation targets for central bank. Government defines target inflation rate, Governor of BoE has, in case, to explain publicly why target was missed.

• By setting an appropriate policy target for the agent, government can correct for a potential policy bias. The marginal incentives for stabilization are, other than when appointing a „conservative“ agent, not distorted.

• Has the „additional benefit“ for government that any failure to reach the target has to be justified by agent, not by government.

• Again, is only credible when agent cannot be overruled by government.

• Another possibility for government to have partial control over the agent‘s policy is to impose a contract on the agent that sanctions the non-achievement of the policy target.

• Should create a stronger incentive for the agent to reach the target.

• May also signal to the private sector that government is really serious about reaching the declared target.

• Example:

  • Contract between central bank and government of New Zealand on inflation target.

  • Manager of state own firms has performance dependent bonus.

  • Ministers are personally held responsible when failing policy target.

• There is a strict technical relationship between policy targets and policy instruments. There must not be too many policy targets, and conflicts and complementarities between targets and instruments have to be taken into account.

• Even a „benevolent“ government faces technical problems in the implementation of policies which imply that targets will usually not be fully reached.

• Strategic interaction between private sector and government creates additional room for policy ineffectiveness.

• Credibility of policy announcements and implementation is crucial. Governments have and use different possibilities to improve credibility, either by employing more or less strict policy rules, or by delegating policy making to independent agents