Is marriage really the answer?

Family First has just released a report on the costs of the breakdown of the family unit. I might not have given it a second glance save that it is written by Dr Patrick Nolan, the more qualified sibling of our Dear Leader.

The thrust of the document, as you might expect, is that costs of having fewer intact marriages are very high. The report points to a bunch of private and social costs, such as increased risks of poverty, mental illness and infant mortality, and tries to put a dollar value on them. It ends up suggesting that the fiscal cost of the reduction in the number of marriages is about $1 billion per year.

I don’t have a sociology background or the knowledge to challenge any of the assertions made by the report, and I trust that Dr Nolan has calculated his costs in as objective a fashion as possible. However, it’s what the report doesn’t monetarise that is most concerning.

The report dwells on the monetary costs of marriage break-up, but leaves out much discussion of the benefits. Marriages tend to break up because of the emotional costs of remaining in them, as Dr Nolan notes. It is very hard to monetarise those costs and the resulting benefits of divorce. For the break-up of marriages to be socially inefficient it must be the case that people are not considering the full costs of the break-up when they make their decision to divorce.

One reason might be the presence of externalities. Perhaps people don’t fully internalise the effects of their break-up on others. They may not fully consider the effect on their children, or the social and emotional cost to those who are close to them. However, those close to them are also likely to suffer by having to deal with the consequences of their sustenance of an unhappy marriage. Whether the costs of that are greater than the costs of dealing with marital dissolution is a tricky question. However, since the report merely details the monetary costs of break-up, excluding the emotional costs and ignoring the counter-factual, it does not help in answering this question.

A second reason may be that people’s expectations of happiness post-divorce do not eventuate. If people tend to systematically underestimate the costs and overestimate the benefits of divorce then there may be a role for a third party. In this case, government and community organisations may be able to increase welfare by discouraging divorce.

So the report, while it may be accurate as far as it goes, does not shed much light on whether discouraging divorce is an appropriate role for government policy. It considers only monetary costs to divorce, which seem likely to be only a small part of the total welfare impact of marital break-up. It also focusses on the present costs rather than comparing these costs to the counter-factual costs of marriages remaining together.

  • “The report dwells on the monetary costs of marriage break-up, but leaves out much discussion of the benefits”

    I see that the objective of the piece is spelled out for us, namely:

    “The objective is not to criticise particular family types but to understand these costs and to support debate on the ways in which they may be lowered”

    So you have no problem with the first point (an understanding of the costs) but do not think that the second part logically follows (that the target should be the lowering of the costs). That is fair enough.

    However look at this another way, understanding the costs, and how they move with social policy is still an objective description – the only issue occurs if the article move to prescribing policy based on these characteristics alone (rather than describing movements). There is only a single recommendation page (pg 30) and that page states that the costs raised provide an issue that should be taken into consideration in policy analysis. In other words the document accepts that it is only giving part of the picture and must be integrated with a more general social welfare function – one that includes the appropriate benefits associated with the decision to divorce.

    Also it is important to think about one of the major reasons why we may be interested in the specific “costs” of marriage breakdown – namely pecuniary externalities stemming from the taxpayer effectively subsidising family separation.

    Although the households involved may be better off if they are allowed to divorce, the fact that society may then have to pay the individuals involved through the DPB and unemployment benefits places a cost on the rest of society – in the end you may end up with “too many” divorces, because people are taking this externality associated with the tax system into account.

    By focusing on individual costs and benefits we miss this, higher level, institutional cost, associated with marriage breakdowns – a factor which the report aims to bring to light.

  • I don’t disagree with the report, as I think i was clear about. However, I think it is misleading to view it in isolation, as you point out. All I tried to do was to set it in context by pointing out the issues which are not addressed by Dr Nolan.

    The use of collective services certainly imposes an externality on everyone and clarifying that cost is useful. However, I think that the decision to provide those welfare services indicates a willingness on the part of society to subsidise individual decisions which result in deprivation.

  • “All I tried to do was to set it in context by pointing out the issues which are not addressed by Dr Nolan”

    Fair enough. However, your discussion gave the impression that these were issues that were relevant to the study he was undertaking – when given the scope of the report (to determine a set of one sided costs) this was not the case. Surely, if all you are doing is placing the study in a broader context then you agree with the quantification applied to a given subset of the analysis required for an ultimate policy decision.

    “However, I think that the decision to provide those welfare services indicates a willingness on the part of society to subsidise individual decisions which result in deprivation.”

    I also have a great belief in that argument – no doubt as we are the two “lefties” on this blog. However,we have to ask three questions here to see if this assumption is appropriate:

    1) Does the public have full information surrounding the costs – if they don’t then this does not necessarily imply that what is actually happening is in the social interest.

    2) Can the public effectively target where they want there spending to go – it may be the case that the public wants to help certain groups, but to do so involves helping other people they don’t want to. As long as the net impact is positive this is fine – however, this factor in conjunction with information problems is more likely to lead to interventionist failure.

    3) Does society as a whole really control policy, or is it controlled by a different group. If a subgroup controls policy then current policy action only represents the preferences of the subgroup – not of society as a whole.

    The answers to these questions will have a great impact on the optimality associated with our welfare policies – just because a certain set of institutions exist does not imply that these institutions are optimal.

  • “Surely, if all you are doing is placing the study in a broader context then you agree with the quantification applied to a given subset of the analysis required for an ultimate policy decision.”

    I don’t have the knowledge or expertise to pass judgment on Dr Nolan’s quantification which is why I didn’t appraise it. I certainly think that the media release arising from the report misrepresented it as a more complete piece of policy analysis than it is. However, that is not likely to be the fault of Dr Nolan himself.

    “just because a certain set of institutions exist does not imply that these institutions are optimal.”

    Certainly. I would prefer to say that, since they exist, they are likely acceptable to the majority of the nation. If the contrary were true then I’d expect a major political party to be campaigning to dismantle those institutions, which isn’t the case.

    The chances that institutions, which have evolved as much as been designed, are optimal seem slim at best.

  • “I certainly think that the media release arising from the report misrepresented it as a more complete piece of policy analysis than it is.”

    Media releases always mis-specify things – the purpose of the document should be defined by the actual document itself, not the value judgments other people attach to it.

    “If the contrary were true then I’d expect a major political party to be campaigning to dismantle those institutions, which isn’t the case.”

    Only if people in society had full information, and if the group controlling the political process has social welfare as their primary target – there is no indication that those assumption hold at all, so the belief that institutions are optimal becomes tenuous in some circumstances.

  • “Media releases always mis-specify things”

    Well, I think that’s a bit extreme. Besides that, the release was done by Family First who commissioned the report. I wouldn’t attribute it to Dr Nolan, but it is clearly associated with the report.

    “Only if people in society had full information, and if the group controlling the political process has social welfare as their primary target – there is no indication that those assumption hold at all”

    Society may not have full information, but is the information imperfection such that they might disapprove of the institutions and yet not know it? I think that’s a bit of a stretch when one considers the attention that redistributive arms of the government get from media and politicians.

    On the second point I agree. I’m not sure that any group controls the political process but, insofar as political parties can be said to do so, it must be true. However, surely their pursuit of electoral success is correlated with their success in improving social welfare.

    The assumptions you state may not hold, but they are not necessary conditions; that is particularly treu when assessing the contention that the institutions are ‘approved of by the majority of the electorate’ rather than the far more stringent test of optimality. I fear that optimal institutions exist only in the minds of academics.

  • “Well, I think that’s a bit extreme”

    As you get to experience more press releases you’ll come to this view 🙂

    “Besides that, the release was done by Family First who commissioned the report”

    So. They applied their value judgments to the report and came up with conclusions that you disagree with – that is fine, the criticism is of the normative assumptions, not of the initial report that they were applied to.

    “Society may not have full information, but is the information imperfection such that they might disapprove of the institutions and yet not know it?”

    If they have poor information about the cost of policies – such that the expected cost is lower than the actual cost, then they will vote for more regulation than they actually want. If people who control the political process have an interest in extra regulation, then they will provide mis-information in order to make this happen.

    “The assumptions you state may not hold, but they are not necessary conditions; that is particularly treu when assessing the contention that the institutions are ‘approved of by the majority of the electorate’ rather than the far more stringent test of optimality”

    Your criticism stems from the fact that you think current policy reveals an implicit preference for that sort of policy – if you can’t even show that institutions are formed to turn peoples votes into the right political institutions how can you make that assumption? It seems like a taller normative assumption than anything in Dr Nolan’s report 😉

  • “They applied their value judgments to the report and came up with conclusions that you disagree with”

    No, I don’t disagree with their judgments; I object to the misleading nature of their release. They represent the report as something it’s not.

    “If people who control the political process have an interest in extra regulation, then they will provide mis-information in order to make this happen.”

    Very likely true, but is there any evidence that it represents a real life situation?

    “if you can’t even show that institutions are formed to turn peoples votes into the right political institutions how can you make that assumption”

    Are institutions not formed by the elected government? Do people not elect the government that they believe will most benefit them? Would you not then expect a correlation between peoples’ desires and the actions of the government? I don’t think that’s a normative judgement; it’s just following incentives.

  • “No, I don’t disagree with their judgments; I object to the misleading nature of their release. They represent the report as something it’s not.”

    Your post is based on the report – you criticise the fact that the report “leaves things out”, without actually touching on what the report is trying to cover. You link to the report, you discuss the report, I’m not even sure you mention the media release – as a result, it would have been hard for me to interpret this as a objection to the media release 🙂

    “Are institutions not formed by the elected government? Do people not elect the government that they believe will most benefit them? Would you not then expect a correlation between peoples’ desires and the actions of the government? I don’t think that’s a normative judgement; it’s just following incentives.”

    If institutions are formed by government, we have to ask how people get into power. Even if we started with a clean slate (such that there was no prior history dependence in institutions which derives the way they evolve – which is a HUGE assumption), if any subgroup is able to shape information in a way to sell their own point of view, then they gain the ability to form institutions in a way that it too their benefit – not societies.

    The ability of groups to use information to put things in there favour, such as Labour has done with the concept of taxing and spending by misrepresenting costs and benefits (they stopped printing productivity measure for many public industries once they cam into power), is part of the description of reality – and it ensures that the choice of government will not always represent peoples underlying desires.

    Also even if there was no information problem – a vote is a discrete thing that happens once every 3, 4, maybe even 5 years. People do not have the ability to reveal there preference in every issue – just some aggregate valuation of the issues avaliable at the time of the election. This imperfection in conjunction with Arrow’s impossibility theorem surely indicates that social preferences are often not revealed by the interests of government.

  • “You link to the report, you discuss the report, I’m not even sure you mention the media release”

    No, I criticised it in the comments. In my post I noted that the report doesn’t provide a complete picture of the policy decision. I did not take issue with the substantive analysis of the report anywhere.

    “This imperfection in conjunction with Arrow’s impossibility theorem surely indicates that social preferences are often not revealed by the interests of government.”

    I think those things imply that social preferences may not be revealed by the interests of government. Perhaps your estimation of how often they aren’t revealed differs from mine. I don’t think that that empirical issues can be decided by reference to possible causes of the difference.

  • “In my post I noted that the report doesn’t provide a complete picture of the policy decision.”

    But if you were only discussing the report why would this matter – when the report explicitly stated it was only looking at one small part of the overall policy decision. The report admitted it was of limited scope – the value of the report came from the substantive analysis of this subsection of the relevant policy elements.

    If you had framed your post in the positive, (eg as well as these costs, another important component for policy analysis would be these benefits) then your discussion would be fine. However, you called the lack of “benefit” discussion “concerning” – this implies that you characterised a problem with the report. However, I think the only way you could find the characterisation “concerning” is if you misunderstood the scope and purpose of the report. A factor I wished to clear up in the comments.

    “Perhaps your estimation of how often they aren’t revealed differs from mine. I don’t think that that empirical issues can be decided by reference to possible causes of the difference”

    I can buy that as a normative difference. However, if you accept that there can be a wedge between what government does and social preference you must admit that this statement:

    “the decision to provide those welfare services indicates a willingness on the part of society to subsidise individual decisions which result in deprivation”

    Is in itself incredibly normative – and as a result cannot be treated as proof that the externality is fully internalised in society. This matters, as it implies that analysis of this externality has value – when with that statement you were implying that the associated externality was not something we should worry about.

  • “However, you called the lack of “benefit” discussion “concerning” – this implies that you characterised a problem with the report.”

    I do think it’s a problem: it’s a problem of omission. I think that a report intended to influence public opinion on government policy should make its limitations clear. The report is very clear on what it says, but equally brief about what other factors might be relevant for the policy decision it seeks to influence.

    i don’t have a problem with the substance of the report, only the framing.

    “with that statement you were implying that the associated externality was not something we should worry about.”

    I wouldn’t go that far. I’d say that people appear relatively happy with the cost of the externality being spread over all taxpayers. I don’t think it is internalised at all.

  • Hi everyone, I’m new to this blog but I really love the content. If anyone is interested in the economics of Asia, you should really check out http://www.asiaecon.org. It’s a great resource for exclusive research, statistics and news.

  • “i don’t have a problem with the substance of the report, only the framing.”

    Hmmm. I wouldn’t have a problem with that, except for the fact that after I read the report I felt that the scope of the report was blindingly clear. It is limited scope is consistently mentioned.

    If the report had to build a case for the other factors (as for this factor it builds a case and then quantifies the elements) that are required to do a complete policy analysis is would be twice as long and would invariably be wrong as it would miss something.

    The sort of document you seem to want does not exist in the consulting world – and I’m loathe to say that it may be a worse type of document than the reports consultants produce given its seeming refusal to admit its own limitations.

    “I wouldn’t go that far. I’d say that people appear relatively happy with the cost of the externality being spread over all taxpayers. I don’t think it is internalised at all.”

    Then we agree here – we were just shifting towards the viewpoint from separate angles. We should save discussion of institutions for a post or something 😛

  • “after I read the report I felt that the scope of the report was blindingly clear”

    Perhaps this depends on the audience. I just re-read the executive summary and it wasn’t until the third page of four that it was mentioned that the costs are gross rather than net. You may expect that an economist would mean it in this way, but if I read ‘costs of divorce’ in a newspaper I would presume they meant that divorce had a net cost to society.

    If I viewed the report as an economic report then I might be OK assuming they were talking about costs the way an economist would. However, since it seems intended to be a policy discussion document, I’m not so sure that such an assumption is warranted.

    Similarly, the estimates are “qualified by the need for further research and debate on the assumptions employed in this paper”. I would be fine with that in an economic paper, but I feel that a discussion paper should at least mention some of the obvious issues that also need to be considered in the policy debate, even in passing.

    The way I read it, it seems like an analysis without context. Perhaps you think that it isn’t the job of the paper to provide a context. Yet, even in academic papers which are intended only for experts in the field, the context and limitations of the research are specified in the introduction.That’s why the lack of context feels like a deliberate omission to me.

  • “That’s why the lack of context feels like a deliberate omission to me.”

    It appears to me that our debate now lies on whether we feel the trade-off between clarity and full descriptive accuracy was made appropriately.

    Any discussion on the trade-off is normative in nature – unless we had some appropriate tools that we could agree on to analyse the issue I am not sure if we will every be able to “agree”.

    My impression was that the report aimed to describe part of the policy Issue – and was clear about the fact that it was only part of the issue. I don’t think we can ask anything more from a report without loosing a substantial amount of clarity – either by making the document too long or confusing the results of the document.

  • OK I`m no economist, but can I jump in here? My impression of the report was that Family First wanted it to be a promotion of marriage as they consider marriage should have fiscal benefits and it is the preferred way of bringing up kids

    but look at this – I`d like your guys’ opinion of this FF calculator. Mine is here

    Although the report – which is titled the fiscal benefits of marriage – does provide a fiscal and social cost of family breakdown, it does not quantify the fiscal benefits of marriage, despite its title, as they are lumped in with – and are no doubt pretty much equal to – the benefits of cohabiting couples – and more than 40 percent of couples aged under 44 are unmarried. But the report discusses decreasing marriage rates,implying that there is a social and economic cost because proportionally fewer people are getting married. But marriage rates are decreasing not just because of cohabitation, they are also decreasing because the parents of a third of this country’s children have no partner. And it is the sole parents in poverty and on benefits who are disproportionately adding to these fiscal costs.
    more here