Home / CONTRIBUTIONS / Time Inconsistency – Time for a rethink?

Time Inconsistency – Time for a rethink?

By GUS DA SILVA 02/10/2014

Pre-commitment rules are often cited as a powerful tool in mitigating time inconsistency – but do they always result in the optimal policy? And is a time consistent policy always preferable?

Bank of england

A favourite literary allegory for economists to illustrate the problem of time inconsistency is the story of Odysseus and the Sirens. Odysseus, wanting to return home safely, orders his men to plug their ears so that they keep the ship to its course and do not stray to the rocks, were they to hear the Siren’s song. Odysseus ties himself to the mast, and orders his men not to obey his orders as they pass the Sirens, recognising he will be behaving irrationally – ensuring a safe passage home.

This is ‘pre-commitment’, a constraint placed upon an actor to ensure adherence to their long-term goals, to the exclusion of choices that may be tempting in the short-run, but potentially costly in the long-run. This commitment to the long-term preference is a state of ‘time consistency’; the tying of Odysseus to the mast is a mechanism to maintain a ‘time consistent’ preference – the desire to return home safely (Strotz, 1955-56). Pre-commitment in practice generally takes the form of institutional rules; these have been advocated as a solution to the time inconsistency problem, starting with Kydland and Prescott’s (1977) Nobel-prize winning paper ‘Rules Rather than Discretion’. These rules aim to solve the problem by limiting the ability of the decision maker to act on tempting short-run incentives.

Time Inconsistency in Monetary Policy

Time inconsistency is a well-established theory with applications in both behavioural economics and monetary policy. Policymakers such as central banks wish to maximise the social objective function – in other words minimising inflation and unemployment. Policymakers thus promise to keep interest rates low, and in considering this policy choice along with other factors, private decision makers respond by adjusting their economic decisions such as hiring, trading, borrowing and spending. Rule based policy reduces unpredictability for these decision makers. Once these resulting decisions have been made, the policymakers’ short-run incentive then changes, the temptation arises to renege on their promise and close gaps in the economy caused by low interest rates.

Perhaps we may, at first, consider discretion to be the best option – it permits flexibility for the central bank should unexpected economic shocks occur. However, there is a problem with this. If the policymaker were to renege on the promise, then their credibility will be damaged – in turn affecting the tendency for private decision makers to believe and act according to its promises. For monetary policy to be effective, the private sector must believe the promise of an extended period of low nominal interest rates – as is currently the case – to prevent large declines in consumption and inflation. This is why even after a deep recession, as in 2008, the central bank may need to credibly commit to keeping the nominal interest rate low for some time even after the crisis has occurred and the economic recovery has strengthened (Nakata, 2014).

For the policymaker, the short-run attraction of acting on the present social objective – closing unemployment and inflation gaps – must be ignored so that they may have credibility to act on, and have their promises believed, during more ‘serious’ future crises. The current social objective is discounted for the indeterminate, yet greater social objective at a later time; sacrifice welfare now, to ensure we do not sacrifice more in the future. This may raise questions for some as to whether we should discount present suffering through a rules-based approach in order to achieve a more optimal outcome.

Time Inconsistency in Foreign Policy

Perhaps less well explored are the policy implications for foreign policy in war. Here, policymakers generally state that the social objective function is the reduction of deaths and a return to peace, both in the short and long run. When this objective appears not to be achievable in the long run, but remains the priority for policymakers, the rational incentive appears to be to renege on the promise to commit to the war and take the time inconsistent option to withdraw. However, the decision is not this simple. Credibility arises again as a means to achieving the social objective function in the long run. In foreign policy, the need to maintain a credible projection of power to those who may present a security threat is an important part of preventing war, by deterring conflict and therefore ensuring the future, ‘greater’ social objective function.

Consider the Vietnam War and the War in Afghanistan. In both conflicts, policymakers stated goals to the public was to achieve the social objective function of restoring peace and stability, and thus made a commitment to defeat the insurgency and remain in the conflict until this could be maintained. When these objectives were found to be unreachable, with the conflicts becoming decades-long ‘quagmires’, withdrawal may seem to be the most rational and optimal response in both cases, by minimising further deaths for a long-run goal that may be simply unattainable.

This, however, still risks damaging the social objective, although to a potentially greater degree. That is, an increased risk of more significant future conflict if the credibility of the decision makers is challenged in the eyes of the insurgents, or other adversaries who take a similar role to the ‘private decision makers’. These actors, it could be suggested, respond accordingly to the policymakers’ decisions. Foreign policy decisions that strengthen credibility may deter these adversaries in the future. Conversely, those that weaken credibility, such as withdrawing, may mean future adversaries are emboldened to undertake further acts of violence.

The seemingly suboptimal yet time consistent policy is to remain in the conflict. In this case, the further deaths of soldiers and civilians in the present are discounted against those in the future. The problem with these situations is that pre commitment rules do not necessarily result in an optimal outcome. Rule-based commitment may even exacerbate violence, if persistent force aimed at reducing conflict actually fosters greater opposition and belligerence.

On the other hand, policy rules may fail due to too wide a degree of discretion. The difficulty lies in being able to predict the decisions of private decision makers in response to policy rules. If private decision makers do not believe or fear the promise or threat, because policymakers themselves are too unpredictable or unreliable, the commitment mechanisms have failed.

As a contemporary example, the conflict in Ukraine could be examined as an example of this failure, and the costs should this occur. In this scenario, NATO may attempt to stick to a rules based, commitment policy, promising to act if Russia instigates further. However, as soon as this is reneged on to some degree, Putin is emboldened to act with further belligerence, weakening the credibility of future commitments (Levy, 2014). New commitment rules are drawn up, such as threats of future sanctions, but as past threats are seen as implausible, the process begins again and risks allowing further escalations in tension.

Rules, Discretion – or ‘Standards’?

The result is a difficult balancing act of rules versus discretion, and the risks associated with both time consistent and inconsistent policy – they may both result in an overall negative social outcome. Kydland and Prescott (1977) determined that discretion was undesirable, as results would be optimal in the short run but inconsistent in the future. In the messy foreign policy world of war and intervention, pre commitment rules can, and have, backfired. In monetary policy, the risks of future shocks are difficult to predict – however, present monetary policy must not unnecessarily damage the current social welfare. In both situations, rules may help ensure ‘greater’ future social benefit, but this has costs in the present: unemployment, in regards to monetary policy; or death of combatants and non-combatants alike in foreign policy. Perhaps we should not always discount the current social objective function for what is usually a speculative future set of benefits.

It is the measurement of potential future risks that determines how much importance we place on ensuring time consistent policy – it is not simply a good in itself. This means we need to ensure policymakers are not completely bound by rules, as these may be made in error, or new situations previously unthought-of may arise; but similarly they must not be permitted too much leeway to make poor, present-biased, decisions. What is needed is a balance: neither complete discretion or rules, but standards, to determine when consistency is optimal, and when inconsistency is permissible (Posner, 2013).

The issue is not of whether we tie Odysseus to the mast, but rather how tight we might tie him.


Kydland, F., & Prescott, E. (1977). Rules Rather than Discretion: The Inconsistency of Optimal Plans. Journal of Political Economy, 473-473. Retrieved from


Levy, P. (2014, May 6). Putin, Ukraine, and Time Inconsistency. Retrieved from http://www.foreignpolicy.com/posts/2014/05/06/putin_ukraine_and_time_inconsistency

Nakata, T. (2014). Reputation and liquidity Traps.  Finance and Economics Discussion Series, Divisions of Research & Statistics and Monetary Affairs, Federal Reserve Board. Retrieved from


Posner, R. (2013). Rules versus Discretion in financial and Other Regulation. Retrieved from


Strotz, R., H. (1955-1956). Myopia and Inconsistency in Dynamic Utility Maximisation. The Review of Economic Studies, 23(3), 165-180. Retrieved from


Check Also

In Conversation with David Halpern Behavioural Insights in policymaking and economic theory

BY OLIVER BARTRUM, 27/11/2016 How can we encourage people to save more? How can we ...

Leave a Reply

Your email address will not be published. Required fields are marked *