This is in fact the first thing that was exciting enough to talk about here! That's why I'm breaking my promise not to post anything until the end of the exams. Quite fittingly, this is exactly the problem I will talk about, the inability of people to keep their promises!
The inability of an actor to credibly commit to a long-term decision, goal or promise arises from the mismatch of short-term and long-term interests, i.e. the time inconsistency problem. (I will draw upon the works of Majone and Rodrik here.) The actor knows what's best for him/her in the long-run and makes a promise. But such a situation arises now that the actor feels compelled to deviate from the optimal long-term strategy. The other actors in the interaction expect this, and the prevailing outcome is suboptimal. The inability of an actors to generate trust on others hurts them before everone else.
The first example Rodrik gives is a democratic country's attitude towards a dictator abroad. The democratic country disapproves of the undemocratic regime and the dictator's illegitimate actions and promises to impose sanctions on the dictator. However, when the possibility of his own people toppling him becomes a reality, the Westerners agree to provide him exile to prevent a civil war. This does not exactly have the effect of discouraging the dictator from being a dictator!
Another example is a government striving to get a multinational company invest in its territory. It provides various incentives to lure the company, but the company isn't sure whether those incentives will be reversed after it undertakes the costly investment. If the government cannot find a way to convince the company, it may well lose a big opportunity that will create much output and employment.
Then there is monetary policy. The government knows that it should stick to a low-inflation policy. (We are assuming there is no independent central bank.) However, a recession (and/or large government debt) makes it very attractive to create surprise inflation. That way, the government will be able to boost output and employment by diminishing real costs. Moreover, it will be able to reduce the real value of its debt. The other players know this and adjust their expectations. Because they expect higher inflation, they bargain for higher wages. In the end the economy ends up having a higher inflation rate than it could have had the government been able to credibly commit to price stability (for a given unemployment level.)
This problem is exacerbated by the fact that elected officials stay in office only for a limited time. The long-term gains of sticking to a policy are after the next election, while the short-term gains from deviation are here and now (the shorter the time horizon in the game, the greater the incentives to cheat.) Moreover, newly elected governments can always renege on their predecessors' promises.
Governments have to come up with ways to tie their own hands. Delegation of powers to an independent, unelected agency with different interests may be the solution. This is the reason behind the creation of independent central banks. Governments give the responsibility of monetary policy to an independent body that doesn't have to care about the next election, but faces great reputational costs from deviating from its optimal long-run policy. However, a national central bank can still be persuaded to respond to a deep recession or unsustainable government debt. Therefore, joining a fixed-exchange rate regime or the Eurozone is an even more credible commitment device. By giving up control over their monetary policies countries attach a new cost to surprise inflation. When they can no longer devalue their currency, higher inflation directly translates into a loss of competitiveness in the world market.
Delegation of powers is not limited to monetary policy. Member states in the EU delegated powers of agenda-setting and law enforcement to the European Commission, which brings them to the European Court of Justice if they violate the Competition Policy. If they didn't tie their own hands this way, it would be politically impossible for them to cut state aid. But then their own exporters would also suffer from state aid abroad.
The credibility-commitment problem, of course, affects our daily lives as well. The grand example is how people never show up for study groups! And I'm sure we all heard of the ham-egg sandwich.