1The article “Covering the Costs of Divorce: The Role of the Family, the State and the Market” analyses the economic risk associated with divorce, and public and private coverage of this risk. The authors mention, furtively but at least three times, that the risk they are discussing applies to unmarried as much as to married couples. And indeed, the various divorce-related costs evoked by the authors also exist for unmarried couples who separate (though the amounts involved may be different). Living standards drop because economies of scale are smaller after separation; human capital is lost because of specialization in domestic labour prior to the break-up (leading to reduced employability and a smaller pension for one of the partners); there are collective costs in terms of child poverty owing to reduced living standards (with a long-term impact in terms of aggregate human capital); there are also costs linked to gender inequality, and procedural expenses.  So if the risks are of a similar nature, one would expect the state to offer the same instruments for covering them, regardless of marital status. And in fact it does so in several respects. For example, welfare benefits for lone parents and people on low incomes, back-to-work incentives, child support guidelines, family provisions in pension entitlements and some family-based tax benefits do not depend on marital status. So why, then, is alimony reserved for married couples only?
2To address this question, we use the same economic reasoning as the article.  If we consider that alimony is intended to cover basic needs, in that the separation penalizes one of the former partners in economic terms, even to the point where he or she (and the child or children) risks falling into poverty, then nothing justifies reserving alimony for married couples alone. We can draw a parallel here with child support, which applies to both unmarried and married parents, since the couple’s break-up in no way affects the parents’ responsibility for their children’s maintenance and education.
3The question is more complex if we see alimony as compensation, and the union as a contract. When the union is formed, or when the partners make decisions later on, they establish a long-term contract under which the partner who, in the early part of their life together, specializes in domestic production (mainly child-rearing) and so invests less in the labour force, will profit from their investment later in life by benefitting from the other partner’s employment income (current or future). In the event of a break-up, the partner who has invested in domestic labour should be indemnified because the union has not lasted long enough for him or her to reap the fruits of their investment.  Can this reasoning be applied regardless of marital status?
Is specialization (and its consequences) specific to married couples?
4From a theoretical standpoint, the answer is certainly “No”. The theory of specialization in the model of family economics devised by Gary Becker makes no hypothesis about marital status. From an empirical and descriptive standpoint, some studies show that such specialization exists in unmarried couples, though to a lesser extent than in married couples (Baxter, 2004; Craig and Mullan, 2010; South and Spitze, 1994). From an empirical and analytical standpoint, however, we know of no studies that have analysed the impact of conjugal status (married versus unmarried) on the human capital (wage rate) of individuals living in couples.
5Instead, the literature studies the effect of being married or unmarried in comparison to being single. The most recent literature on this subject shows that there is a (slight) marriage premium for men and an even slighter (or nonexistent) premium for men in unmarried couples (Barg and Beblo, 2009; Cohen, 2002; Datta Gupta and Smith, 2002; Datta Gupta et al., 2007; Dougherty, 2006; Killewald and Gough, 2013; Mamum, 2012; Pollmann-Schult, 2011; Taylor and Bardasi, 2008).  A small marriage premium has also been observed for childless women, but no statistically significant effect of being in a couple for mothers (Datta Gupta and Smith, 2002; Dougherty, 2006; Killewald and Gough, 2013). By contrast, the literature quite clearly agrees that motherhood carries a penalty (for an overview, see Jeandidier and Lim, 2015). We can draw two conclusions from these studies. Marital status does not seem to be a strong determinant of specialization or of the consequences of specialization; but whichever the type of union, if there are children, the consequences of motherhood are undeniable. So from this point of view, where the couple has children it does not make sense to restrict alimony to married couples alone. This is precisely the recent conclusion of Canada’s consultative committee on family law (2015), a committee of legal experts who argue in favour of a parental compensatory allowance independent of marital status. 
Does the nature of the contract differ by marital status?
6From the standpoint of law, legal experts would say that marriage being an institution,  it provides sufficient grounds for treating the two kinds of couples differently. From the economic standpoint, the question is less clear-cut. It seems to us that whatever the marital status, there is an implicit, unwritten contract; the partners trust each other and legal marriage changes nothing. The career choices the couple make, with the promises that go with them, are not explicitly a part of the marriage institution. But it could be objected that alimony is an aspect of the marriage institution, that the implicit contract takes it into account, and therefore that, all other things being equal, the implicit contract is not the same for married and unmarried couples. This seems to be the hypothesis developed in Landes’ theoretical model (1978): in marriage, the spouses adjust their career choices according to the payoff they expect from alimony in the event of a break-up (and hence for an unmarried couple the payoff is certain to be nil). However, the pertinence of this argument is doubtful, at least in the case of France: do couples really make their career and fertility decisions in full knowledge of articles 270 and 271 of the Civil Code? Who can reasonably estimate the amount of alimony far ahead of time, when there are no official guidelines, and when judges and lawyers themselves admit to being ill-equipped to set the appropriate rates? Clearly, it is fair to doubt the effectiveness of alimony as an ex ante incentive. We can therefore consider that the implicit contract is of the same kind whatever the marital status.
7We thus find few arguments in favour of maintaining this unequal treatment between married and unmarried couples. Regardless of marital status, one of the partners, usually the woman, is making an investment when they stop paid work to look after the children  – despite the current (slow) behavioural trend towards less specialization of labour between partners. Alimony should be extended to unmarried couples so as to acknowledge that investment. This would also give added strength to the policy of gender equality.
BETA, CNRS and Université de Lorraine.
Even though there is no divorce procedure, there may be legal expenses involved in soliciting a judge and a lawyer to set the child support amount, or notary’s fees for settling property issues, etc.
We do not approach the issue from a legal perspective, since this is not the authors’ angle of attack.
As the authors explain, this compensation mechanism also constitutes an ex ante incentive to stay together, or to make career choices according to the likely level of compensation.
The difference in the premium for men in marriages and for those in non-marital unions could be interpreted as a consequence of lesser specialization in unmarried couples, but it should be stressed that, to our knowledge, no study has explicitly tested the statistical significance of this difference.
This would be in addition to child support, which follows a different logic: child support divides the current and future costs of the children between the parents, whereas alimony compensates one of the partners for the investment they have made in the children’s past education.
From the legal point of view, marriage is not a contract but an institution, because the parties cannot change the rules of marriage, which apply to everyone in the same way.
We suggest limiting alimony exclusively to parents because (a) the career choices deserving of compensation are mainly connected with motherhood, and (b) parenthood confirms implicitly– in a certain manner, although this is debatable and needs to be examined further – the veracity of the union (while the cohabitation certificate and civil union do so explicitly).