Love and Information: The Return

Note: A slightly more technical discussion of some Industrial Organization applications to sociology.

The hold-up problem is caused by asset specificity and absence of a perfect contract in a market transaction. These two elements are so broad and commonplace that they are likely to be found in almost all aspects of life. One of these is the realm of romantic relationships. When two individuals begin a relationship, usually through a “dating” process, this is the first instance of several future decisions that have to be made. These decisions can be analyzed using some very conventional tools of economics.

In a utility-maximizing framework, it is understood that the individuals are making a decision whether to enter a relationship with the other person on a utility-maximizing basis (i.e. all relatively new relationships are understood to be revealed preferred to other available relationships). One could think of this new organization between the two individuals as a firm committed to producing several different outputs: Joint leisure, children, economies of scale (sharing apartments, cars, etc.) and even joint economic output (through division of labor).

One of the things that makes this type of organization special is that, throughout its conception, it is not (usually) governed by an explicit contract. Rather, it is until much later, when the partnership has been judged by both parties to be sufficiently promising, that a formal contractual agreement (e.g. marriage) usually takes place. It is interesting to note that even this contract is much further from theoretical “perfection” than any standard business analog. For its maturing stages then, a relationship will be guided by an informal contract adopted from social norms. For convenience, we will not describe its clauses, trusting that they are familiar to the reader (e.g. sexual exclusivity, joint investment of time and effort, etc.).

It is these beginning stages of a romantic relationship that are most prone to under-investment due to a potential hold-up problem. First, consider the relationship-specific investment in knowledge of the counterpart: For both persons involved, investing in knowledge of the other person’s attributes, likes and dislikes is an essential input for production of joint leisure. This investment, however, is virtually useless (and perhaps even harmful!) in any other potential relationship the person might have.

Note also that the cost of this investment is increased further by the fact that a person’s probability of finding a suitable mate is (generally) decreasing with respect to time (due to a decreasing pool of single potential mates, or decreasing physical attractiveness, for example). Once characterized in this fashion, the situation is that of a classic hold-up problem: Due to the uncertainty of a relationship reaching a point of stability (i.e. marriage or any other barriers-to-exit scenario) in which asset-specific investments will pay off, the realized level of investment that must be sunk previous to commitment is sub-optimal.

In the standard economic theory, we see that it is usually the party which must sink the investment that will stand to lose (and which under-invests as a consequence). In this particular case, since both parties must sink some investment, it is the party whose opportunity cost of investment is the greatest that stands the most to lose. This would suggest that this is precisely the party that would propose a binding contract in order to safely sink the investment, or ask for a lump-sum transfer to compensate for the asymmetry. Since the latter case is trickier to analyze in this context, let us stick with the former.

One instance of this is the added opportunity cost of investment for females due to a smaller window for reproduction. Assuming that both parties in a relationship value parenthood to the same (positive) degree, it will be the female who stands the most to lose from a long “going-out” process that fails to come to fruition. A similar, if less politically incorrect, example is that of a relationship between a high-level executive and an individual living comfortably off an endowment (without any labor involved). One would expect, holding all else constant, that the executive’s higher cost of time would lead to a “pressuring” towards an explicit agreement for mutual investment.

Consider now an additional effect of the investment in knowledge of the counterpart (i.e. besides production of joint leisure): Additional information regarding the expected outcome of the relationship in the long term. Thus, agents invest in “getting to know” the other person not only because it improves the mutual experience, but because it allows them to (eventually) decide if the other person represents an inter-temporal utility-maximizing choice of mate.

It then follows that the impact of this hold-up problem is not only a loss of potential joint leisure production throughout the “dating” period, but also that the decision whether or not to sink investments with a particular individual might be made with sub-optimal information. In general, this would imply a greater chance of choosing to sink an investment when it should have been avoided (e.g. marrying a then undisclosed schizophrenic) or not sinking when it should have been made. Since it is difficult to appraise the cost of the first scenario, the millions of dollars spent in divorce settlements in Canada alone (72,000 divorces in 2004 with costs ranging from $540-$18,900) might help gauge the impact of these suboptimal decisions.

It is difficult to suggest an additional contractual resolution to this hold-up problem since such a solution already exists. The entire institution of marriage is a contractual resolution which raises barriers to exit and reduces the probability of wasted ex-ante investments. Other, less rigid contractual resolutions exist: The engagement ring is an intermediate contract between the beginning and “stable” stages. In some countries, such as Mexico, there are specific laws (albeit old-fashioned ones) that give a woman the right to sue an ex-boyfriend on account of “lost time” if he terminates a long relationship that never made it to marriage. These kinds of intermediate agreements (whether public or private) could “ease” the individuals along their investment paths towards the optimal level.

The most difficult part of this problem seems to rest mainly on the difficulty of extracting accurate information regarding the counterpart’s behavior and reliability in the long term. Given what we know about the effects of uncertainty on investment levels, even perfectly designed intermediate contracts are not likely to solve this issue.


  1. MK

    This is an excellent analysis of the holdup problem applied to romantic relationships. This topic has so many features and layers that it has prompted me to write about another aspect of it that you did not cover in your essay.

    Collecting information about someone is relationship-specific investment without guaranteed return, and the hold-up problem predicts that there will usually be underinvestment in getting to know a particular individual before deciding whether to allocate more time and effort for this person (i.e., dating). However, the investment in searching for a suitable mate is necessary in order to start a relationship, and to avoid the outcome of making a choice that turns out to be different from expectations. For this reason, we may conjecture that risk-averse agents may actually over-invest in searching and evaluating potential mates before making a decision. This could explain why some dating periods are so long and why are some people are still single at a late age.

    The difficulty is not that individuals are disinclined to invest in getting to know potential mates (a necessary investment) but because investing in specific persons is costly, they must rely on readily available (observable) information and prior beliefs in selecting a preliminary set of candidates, based on criteria such as desirable traits, mutual interest and likelihood of compatibility. Given that many aspirants simultaneously compete for the attention of your most desired partners, one strategy is to engage in sending the most attractive (and credible) signals. This solution is an imperfect one: although sending truthful signals could mitigate the collective problem of finding the best match for everyone, it is in each person’s best interest to send the signals that best serve him/her. The rationale for misrepresentation–or failure to disclose–your type is simple: your most desired mate may not prefer you as a mate if he/she were to have full information. For example, some people may not be looking for a long-term relationship or may be in some relationship and evaluating alternatives, and since this reduces their attractiveness to the other party and hence could reduce their available choices, are reluctant to reveal their relationship status or preference.

    The existence of individuals unwilling to reveal truthful and complete information is an issue for those who want to send truthful signals to do so in a manner that appear credible to their uninformed targets. In addition, the costs of (selective) signaling–which include identifying a match who appears sufficiently promising, and determining the appropriate information to communicate to win his/her favor–reduce the incentives for investing in signaling your interest and personal attributes. These exacerbate the difficulty for the uninformed party to extract useful information and making the correct inferences from available cues about the candidate partner. The problem is as much for the uninformed party to make the right amount of investment in knowledge of the prospective partner as it is for the informed party to send informative and reliable signals. Indeed, the match-finding process is fraught with incentive and informational problems, resulting in suboptimal matches overall.

    • Daniel Silveyra


      I thought about what you wrote for a long while. That means “thank you” in a convoluted way. As you said, its a pretty complicated issue, and I think that the information asymmetry side of things is very relevant. Let me elaborate:

      In terms of the information-collecting, it seems that we as a species have somehow evolved to enjoy the process of getting to know potential dates, so the cost of dating isn’t prohibitively high. Additionally, when considering long dating periods that do not resolve into “permanent” relationships, I think one must consider the implicit message (i.e. if your boyfriend hasn’t proposed in 5 years, you might do better in looking for another one).

      I like the fact that you introduce competition into the framework – someone suggested that I write up a general equilibrium model for this a few weeks ago. In your proposed scenario, it really turns out to be a “cheap talk” game, where all aspirants are saying what they believe the potential mate wants to hear.

      However, you seem to be operating under the simplifying assumption that preferences are homogeneous (” one strategy is to engage in sending the most attractive (and credible) signals”), i.e. that the same things are attractive for all aspirants. In this case, the “good” or matching candidates send the same signal as the “bad” or poorly matching ones, since it costs nothing to learn what this signal is.

      If one removes this (somewhat unrealistic) assumption, then it turns out that potential mates must invest in getting to know the specific tastes of the other, and thus there exists a sort of filter. I agree that the filter is necessarily imperfect, however.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s