Allan Drazen. 2000. Political Economy in Macroeconomics

Allan Drazen. 2000. "The Time-Consistency Problem" and "Laws, Institutions, and Delegated Authority," in Alan Drazen, Political Economy in Macroeconomics (Princeton, NJ: Princeton University Press): 101-165. Chapters 4 and 5.

Chapter 4: This chapter just gives examples of time-inconsistency in policy choices that might arise. They all arise from heterogeneous preferences/conflicts of interests. 

Terms:

1. Time-inconsistency is said to arise if, though nothing has ostensibly changed, the policy chosen for time t+s chosen at time t is different from the policy chosen for time t+s at time t+s. 

I. Introduction

• The puzzle: why does time-inconsistency arise if the fundamental characteristics of the policymaking environment does not appear to have changed?  
• A conflict of interests of some sort is necessary for time inconsistency to arise.
• Time inconsistent policy is interesting when it is chosen to maximized the welfare of those who are misled.

II. A Simple Model of Capital Taxation

• There are two time periods. 
• In the first period:
• The government announces the tax rates it will implement in period two.
• The individual being taxed has an exogenous endowment and chooses consumption level and capital accumulation to be used in the second period
• In the second period:
• The government implements a tax rate on the capital the individual saved in the first period and the labor the individual gives in the second period.
• The individual gets the payoff from government spending and from consumption, which are functions of the capital saved from the first period and the labor the individual supplies in the second period.
• Time-inconsistent solution: the tax policy the government announces for time period 2 while in period 1 is different from the tax policy the government actually implements at time period 2. Occurs whenever ex post capital elasticity is less than the ex ante elasticity.
• In the first period, the government will announce a tax vector with a low rate to encourage capital accumulation.
• In the second period, the government will carry out a tax that is different from the one announced in the first period and there is nothing people can do about it because they can't really change the capital supply anymore.
• Precommitment solution: the tax policy the government would announce in the first period when it has a mechanism to commit to it and not reoptimize in the second period.
• The precommitment solution is the same as the time-consistent solution that results when individuals take government preferences into account when they choose their income allocations and when governments take into account individual preferences into account when they choose their tax policies.

III. Explaining Time-Inconsistency in the Model of Capital Taxation

• People cannot operate for their own good when they are subject to pre-existing constraints or distortions.
• Sequential policymaking is a necessary but not a sufficient condition for the possibility of time consistency to arise. 
• What is essential to the phenomenon of time inconsistency is conflict of interests in the second time period (ex post heterogeneity).
• The dependence of utility on aggregate allocations induces a source of conflict among agents, which is crucial for the possibility of a time-inconsistency problem.

IV. A Basic Model of Monetary Policy

• The policymaker chooses optimal inflation taking expected inflation as given. 
• The conflict of interests which lie behind the possibility of time inconsistency:
• Mirroring the capital tax problem, heterogeneity of interests in a representative agent model results in conflict of interests. Each individual wants to minimize the error of his own forecast of future inflation, but would like everyone else to under-predict inflation so that the economy-wide average prediction implies low unemployment.
• Conflict not in the capital tax example: conflict of interests between policymakers with different objectives, reflecting perhaps a conflict of interests between the different constituencies they represent. Conflict occurs when the natural rate of unemployment that the fiscal authority finds optimal is not the same as the natural rate that the monetary authority finds optimal; the fiscal authority has an incentive to increase economic activity and thus drive up short-term inflation.

V. Equilibrium Solutions for All Models

• Optimum is achieved when the policymaker is led at time t+s to carry out the policy announced at t, rather than some other policy, and it is "common knowledge" that he will indeed carry out the policy.

VI. Commitment vs. Flexibility

• There might be gains to ensuring commitment, but in the real world, unforeseen and unforecastable events occur so that the optimal policy at time t+s cannot always be identified at time t. 
• Escape clauses allow for commitment and flexibility.

Chapter 5: This chapter provides solutions to time-inconsistency problems. This chapter concentrates on how the policymaking environment can make policy credible, that is, how institutions or the creation of external circumstances (broadly defined) can lead to the expectation that announced policies will be carried out.

I. Introduction

• When policymaking is viewed as a sequence of decisions, so that the government can reoptimize at every point, the problem of time consistency can be viewed as reflecting changes in incentives over time. Time t decisions lead to an evolution of state variables which give a policymaker the incentive to deviate at time t+s from his previously optimal policy.
• Example: the decision to impose taxes on capital in a time-inconsistent way reflects the accumulation of capital, an accumulation that was induced by the government's previous policies. Hence, with the policymaker's narrowly defined objective function unchanged over time, time inconsistency may be thought of as due to change in the environment brought about by the policymaker himself.
• Time inconsistency can be avoided if a policymaker at time t can choose policy in such a way that state variables at time t+s imply that it is optimal not to deviate at t+s from the previously optimal policy. 
• One way to make current policy credible is by building a reputation by engendering the expectation that certain policies will be followed in the future on the basis of actions that have been observed in the past. 

II. Laws, Constitutions, and Social Contracts

• There are important differences between promises which have no legal backing and laws (including widely accepted norms) in analyzing solutions to the time inconsistency problem.
1. Laws have penalties attached to them so that there are explicit costs to breaking the law. Similarly, social norms have recognized costs associated with not following them.
2. Explicit laws or widely recognized social norms make noncompliance more visible and hence more costly. 
• Since laws make policies credible only to the extent that the penalties which enforce the laws are themselves credible, enhancing credibility depends on choosing the optimum structure of penalties to do this. 
• Laws (and institutions more generally) can enhance credibility by raising the cost and lowering the benefit from deviating from a given policy. 
• Effective commitment follows from the extreme difficulty in changing a law once it is given constitutional status.
• Constitutions can make policy more credible because it does the following:
• Restrict government's use of authority. 
• Set out the basic processes of policymaking—laws about how collective choices should be made.
• Treat issues that are more fundamental than others, such as basic rights of liberties.
• Provide stringent amendment procedures than other laws.
• Unwritten agreements that have force because they are generally agreed upon go by several names: social contracts, social conventions, social norms.
• Social norm - a pattern of behavior that is customary, expected, and self-enforcing. 

III. Delegation of Authority

• Delegation from a principal (the government) to an agent (the agency/authority) might occur for the following reasons:
•  The agent may have greater expertise and experience regarding a policy area.
• Governments are required to handle a large number of issues, each of which may be extremely complex, making it impossible for a single policymaker to make all decisions. The number and complexity of issues makes delegation essential.
• The principal and agent can sign an incentive contract to eliminate agent bias in policy choices and ensure optimal outcomes. The contract institutionalizes the incentives for compliance; the cost of changing an institutional structure is higher than changing a policy in itself.

IV. Fiscal Structures for Time Consistency

• A government can bequeath to its successor government a specific debt structure, such as setting maturing debt in each period equal to tax revenue net of government spending, to eliminate the incentive for its successor to change tax rates and thus try to reduce its debt obligations.