'Marriage' in the 21st century: How do we measure relationship formation?

This post deals with some thought that I have been mulling over for a few years now, culminating ultimately in a chapter in the book Young People's Development and the Great Recession:
Uncertain Transitions and Precarious Futures which is launched today (and is available here and here if you are looking for a stocking filler for that special someone in your life).

Essentially the idea is that while traditionally, measuring the start of a union was relatively easy since it was demarcated by a formally recognised institution, marriage. Obviously this is no longer the case- a large proportion of unions are now cohabiting, even when this union is permanent and lifelong, and for all intents and purposes marriage but not formally recognised. This leaves us with something of a problem when trying to measure these unions: relying on indicators such as whether a couple is married or not is obviously not sufficient. If we consider the example below from two of the latent classes from Perelli-Harris and Lyons-Amos 2015: lines in the figure represent the partnership state at a particular age: green represents direct marriage, red marriage where the couple lived together before being married and orange represents cohabitation

Source: Perelli-Harris and Lyons-Amos 2015 pp. 160

As we can see, for the vast majority of the lifecourse, women in both of these classes are married to their partner. The only real difference occurs around age 20: in class 3 women live with their partner before marrying later, while in class 1 they do not. It's not surprising to note that class 3 has effectively replaced class 1 as the form by which life long and early marriage is exercised, at least in the West, as cohabitation prior to marriage became an institutionalised part of marriage (e.g. Cherlin 2010 The Marriage Go Round).

It's this institutionalisation of cohabitation prior to marriage that causes our measurement problem: how do we distinguish between unions which are effectively lifelong unions which will be formalised later- or not at all- from those which are shorter term unions? The idea might be that we look at other indicators which are correlated with permanent union formation which is not marital. This borrows heavily from the idea presented in Holland (2011)- who uses housing purchase as a marker of stability- and some of my other work  which leads the argument: union formation is no longer predicated on a single formal event (marriage) but rather a series of interrelated events with no certain ordering, but which occurring at about the same time in the lifecourse mark a more permanent union state.

Clearly the interrelation of these events will tend to change over time: for instance the waning importance of marriage and well as much lower importance in some contexts means that these correlations will tend to vary across space and time. The focus of the book chapter was to look at this, but also to examine whether certain seminal events (such as the Great Recession) were disruptive and invoke sea changes in the relationship between both demographic and other life course transitions- or merely resulted in postponement of all lifecourse events. The latter is generally true (see excerpt from the book below): both demographic as well as labour force events tend to occur later to women who make their key transitions post recession as we can see below

Estimated probit coefficients for the effect of recession on transition intensities 

Notes: Coefficients are probit coefficents for the effect of the Great Recesion on transition rates: more negative coefficients means slower/later events. Error bars denote 95% credible intervals

Source: Lyons-Amos and Schoon 2017 in Schoon and Bynner 2017 pp. 307

Generally then, the onset of the Great Recession in 2008 tended to retard not only labour market transitions such as getting a job (as would be expected), but also demographic transitions like living with a partner and having children (this latter point I have explored elsewhere in this blog and in publication). This adds some weight to the idea that a broader set of correlated indicators can be used to measure union formation. That said, there is also evidence that the recession changed the interrelation between these indicators: the figure below looks at the covariance between the timing of forming a coresidential relationship and having a child. Although weakly powered, there is some evidence that there is a change in the interrelationship here: post recession unions formed are less likely to result in childbearing at about the same time

Estimated covariance between timing of first child and forming first co-residential union 

Notes: Coefficients are covariance for the effect of the Great Recession on transition rates: more positive  covariance means that union formation and childbearing are more likely to occur at the same lifecourse stage  

Source: Lyons-Amos and Schoon 2017 in Schoon and Bynner 2017 pp. 313

There are probably a few take homes from this then. Firstly, measuring union formation  by single indicators is probably no longer sufficient given the complexity and heterogeneity of union behaviour, and incorporating other demographic indicators such as childbearing or house purchase or even those which we might traditionally think of as belonging in the realms of other social sciences might give us better delineation between more and less permanent relationships. Second, however, the precise relationship between these indicators is not fixed and will not get us back the sort of neat cut-off that single indicators such as marriage did in the past, due to the fact that interrelations will tend to vary across country context, across generations, and as we see in the book chapter, in response to large external shocks. 

Young People's Development and the Great Recession

Uncertain Transitions and Precarious Futures

Is marriage becoming more stable?

There was some minor interest in marriage and cohabitation this week when the ONS published statistics on the rate of marital divorce in the UK, which has continued to fall. This was picked up in the media, for example by the BBC


Divorce rate in England and Wales, divorces per 1000 married

Source: ONS via BBC 

Many of the headline claims are accurate; the rate of divorce continues to decline, and marriage is more stable than cohabitation. However, the major issue is that much of this is often in phrasing the effect as being a causal one, for example:

the declining divorce rate suggests that marriage itself is becoming more stable.

This is only one possible explanation. Another is that marital dissolution rates are partially reflecting selection processes, and these processes vary over time.

There is a reasonable history of finding selection processes in marital formation and subsequent dissolution (here's a classic). The exact manner of this selection will vary: as suggest in the Lillard et al article some of it will be endogeneity: couples who are less confident in their relationship will either cohabit first as a trial marriage before committing, or will continue to cohabit without translating their cohabitation into marriage. As such  marriage will tend to have lower rates of dissolution because they are the relationship which were less likely to fail in the first place. Another route of selection is that marriage is increasingly the preserve of relatively advantages couples: socially advantage seems to depress union dissolution regardless of whether the union is a marital one or a cohabiting one- although the marital advantage remains across all socioeconomic strata.

In either case, the fact that marriage is a select process means that interpreting declines in divorce as being due to the changing nature of marriage ignores the fact that the type of people who marry, and the intensity of selection effects into marriage, will also influence the rate of marital dissolution. It's undoubtedly true that there have been changes to the institution of marriage across time and some of this has influenced dissolution rates. However, ignoring changes in the composition of the married population- the characteristics of married people- is to miss out an important explanatory factor.

Research publication: First births and the Great Recession in the UK

The research publication Differential responses in first birth behaviour to economic recession in the United Kingdom was finally online published today in Journal of Biosocial Science, and can be found here (£).

I'm not going to go through the findings of the paper in detail in this post, but more concentrating on some things that didn't make it into the paper- namely the sort of claims that can be made from natural experiment type of events that the Great Recession represents.

The Great Recession was really fortunate in terms of social science study design (less so in terms of other consequences) as it makes it relatively easy to make causal inferences. There are two key features that make this design nice:

1. The recession was relatively unexpected. Although there had been some instability in financial markets in the upcoming months in terms of US sub-prime and things like BNP, the actual crash itself came as something of a surprise (most recessions are, relatively few populations have repeated warnings about potential economic slowdowns in the medium and long term but still decide to plough ahead anyway). As such, we don't get anticipatory effects that we get where such a 'recession discontinuity' (here all week) is predicted since policy reforms tend to be announced in advance- although there are notable exceptions.
2. The recession, as a cross-national macro-economic event, was exogenous with respect to the individuals it affected. 

As such, the recession is a reasonably nice natural experiment. There are some limitation in terms of what we can claim that said: since this is not a true discontinuity (there is a reason the pun didn't make it into the title) but an economic shock, the causal effect in terms of fertility behaviour should be limited to a relatively short period surrounding the crash before we start to get into problems of what is the effect of the recession and what are the effects of the policy responses to the recession. We didn't end up doing this in the paper due to sample size and because we wanted to make inferences about older women to capture cohort effect: others have although they don't examine the heterogeneous effects of the recession that we do. The paper is therefore capturing effects on fertility beyond those of the natural experiment, although it was probably worth doing so to get an idea of the various effects on different social groups.

DDRedging up the past: historic posts and marital institutions.

This article was published in Demographic Research today, which resonates with the themes of previous posts. There has naturally been previous research on a similar theme, but this paper is nice (from my perspectives) since it is an explicit comparison between the East and West, and also examines the interaction between social norms and behaviour.

On the normative foundations of marriage and cohabitation: Results from group discussions in eastern and western Germany

 Klaener and Knabe

Abstract

Background: Since the 1960s the inclination to get married has been declining in almost all western industrialised countries. Partnership arrangements have become more diverse and the share of cohabiting couples and nonmarital births has been increasing. Yet there are substantial regional differences in marriage and childbearing patterns, and the differences between eastern and western Germany are especially striking. We assume those differences can be partially explained by social norms and different attitudes towards marriage, cohabitation, and childbearing.

Objective: We explore the views and values young people in eastern and western Germany hold about marriage, cohabitation, and childbearing and how those views relate to individual life experiences. We also examine different social norms and contexts in both parts of the country.

Methods: We analyse data from qualitative focus group interviews conducted in in Rostock (eastern Germany) and Lübeck (western Germany) with women and men aged 25−40.

Results: Our findings indicate that there are normative differences between eastern and western Germans in their attitudes towards partnership, marriage, and family formation that can be traced back to the social and political conditions prior to German unification in 1990. The harmonisation of family laws and policies across Germany after 1990 did not automatically lead to a convergence in the norms and behaviours of the people living in these two regions.

 There are a few interesting points here, for instance looking at the effect of legal requirements on certain demographic behaviours: the pointlessness of marriage in the DDR now that some requirements of the Socialiast state no longer pertain is noted:

What marriage meant for our parents doesn’t apply to us today. […] I know that my parents got married because they were having a child, and would then get a flat. That’s not the case anymore. Today I can choose my own flat. [...] that’s the way it was in the GDR. There was the credit for couples with a large number of children, but you only got it if you were married. That was a real status, being married […] To be someone, you had to be married. And that’s been lost today

 Klaener and Knabe (2017) pp. 1651

This is a highly plausible case for a causal effect of an institution on demographic behaviour that I have been skeptical about previously (here and here and somewhat here). While the differences can't be fully attributed to differences in regimes pre-1990 (we saw differences within the German Empire) there does seem to be something close to an effect: the exact identification strategy needs some more exploration that said.

Conference presentation: Discontinuation and selection: using cure models to evaluate the contribution of selection to sectoral differences in modern method contraceptive discontinuation

This is a blog post based on the presentation I'm giving at PAA this afternoon.

Contraceptive discontinuation is a key indicator of the quality of care provided within a family planning programme. Comparison of headline rates of contraceptive discontinuation can be problematic due to differential prevalences of ‘cured fractions’- women who have a zero probability of discontinuation. These women are heavily concentrated among long term contraceptive method dispensed by public sector providers, and comparison of discontinuation rates without accounting for cured fractions can be misleading. This analysis therefore uses cure models to evaluate contraceptive discontinuation using DHS calendar data and a 7 countries in sub-Saharan Africa. Results indicate that a substantial part of slower rates of discontinuation among long term modern method users can be explained by cured fractions. Conditioning on this, discontinuation rates are higher among users of these methods, potentially indicating a greater rate of side-effect related discontinuation and implications for care provided by public sector providers. 

The fundamental technical piece of understanding we need here is the idea of a 'cure model,' aka a split or two population model. The basic idea is that in standard event history models, we are looking at the rate that a particular event occurs over time. The problem being, that under our two population framework, some of the people we observe are not truly at risk of the event occurring. This is common under circumstances where we are looking at the recurrence of a disease, but we 'cure' a certain proportion of patients (hence the 'cure' model). Under the circumstances we are looking at, contraceptive discontinuation, we can think of women as being 'cured' from contraceptive discontinuation under circumstances where they are

1.  Are practising stopping behaviour (so have no risk of contraceptive abandonment)
2.  Are using a reliable contraceptive diligently (so have no risk of contraceptive failure)
3. Are happy with their current method (so have no risk of method switching) 

This leads us to an analysis where using a definition of contraceptive discontinuation consistent with the DHS that is vulnerable to bias, due to the presence of women fulfilling these criteria. The specific bias we are talking about is illustrated below: we are including women in the risk set who have no risk of experiencing the event, are will continue within our population until the end of time

Why this matters is because the distribution of the cured women is likely to be correlated with certain variables we might be interested in- for instance, there may be variation across country or between different providers. Where we are using continuation rates as an indicator of quality for family planning programmes (Jain 1989, Bruce 1990) this is problematic: some of the 'good' performance is merely reflecting variation in terms of the women served and our comparison is subject to a selection bias.

This is probably the most obvious point of attack for this analysis- particularly for contraceptive switching: it's easy circumstances where a woman changes to a more suitable method. I would, however, whether this really detracts from the idea that there is some degree of programmatic failure involved in that decision. Why was the preferred method not available in the first place? If women change method a lot to a string of 'preferred' methods, is contraceptive counselling really adequate? Besides, even if the change really is desirable, there are still positive risks in the contraceptive change relevant to programme evaluation, such as an elevated risk of accidental pregnancy, regardless of whether the switch is desirable in terms of user satisfaction.

The results of the analysis using the DHS definition of contraceptive discontinuation is presented below. We are making cross-national comparisons here, based on a country selection described in other analyses and using the contraceptive calendar, one of the more reliable means of collecting contraceptive use. Results are presented below, comparing the hazard of contraceptive discontinuation for long term methods (IUD, Male sterilisation, Female sterilisation, Implant) to all other methods for both usual event history and cure models.

Looking at the column for the hazard model, we can see coefficient which are generally below 1 for most countries. This indicates that in most setting, users of of long term contraceptives are less likely to discontinue contraceptive use that women using a short term method (this is a fairly intuitive result). In terms of a naive policy recommendation, this would tend to point us toward encouraging the use of long term methods: where lower rates of contraceptive discontinuation are desirable from a quality perspective this would tend to be a means of lower the rate at which women stop using contraceptive.

However, the column on the right, which accounts for our cured fraction has a slightly more mixed result. Of the 5 countries which had significant coefficients below 1, only 2 remain (Burkina Faso and Malawi). In contrast, we now have 4 coefficients which are above 1, indicating that, accounting for our selection effect, women who use a long term method are more likely to stop using that method than short term method users. Our policy recommendation here is radically different: our call would be to improve the quality of contraceptive counselling for long term method users to reduce their inflating discontinuation rates. This is particularly relevant where there is a greater propensity for women to accept a long term methods, such as in the public sector (Campbell et al 2015). Moreover, it reinforces the conclusions I have made in previous posts

selection effects exist within family planning and that failing to account for these can lead policy makers to overplay their hand when advocating certain solutions. If we want to use genuine evidence based policy we need to think about the nature of behaviours we are generating evidence about: simple regression may not be adequate.

Are imposed institutions exogenous?

This is post is something of a mirror image of previous posts that I have made, where I have talked about the potential endogeneity of demographic behaviours and the institutions within which they occur (see here and here). So the task here is to try and find a circumstance whereby we might be able to identify a set of social institution which were truly exogenous to the populations on which they were imposed- this would allow us to start to get at estimating causal effects of policy on demography. It also means that I get to mention Prussia, the history of which is fascinating for me but possibly boring for everyone else.

Socialism: an exogenous social institution

A natural place to look for these type of institutions would be the European countries of the Soviet bloc, and in particular the Deutsche Demokratische Republik. The major demographic difference here is in terms of non-marital childbearing: non-marital childbearing is much more common the East than the West, a behaviour mirroring the attitudes of the citizens of those regions (Perelli-Harris and Bernardi 2015)

Non marital births within Germany 2009

Source: Klüsener and Goldstein 2012

Legend denotes share of births to unmarried mothers

The argument for why the Socialist institutions we saw in these countries is based on the logic that Socialism was largely imposed on these countries by invading Soviet armies- there was no Miracle of the House of Brandenburg this time (you were warned). As such, the imposition of the Socialist institutions post-war should be exogenous to the underlying characteristics of the country in question. (You cannot make the same claims for the USSR incidentally, the fact that Socialism emerged here following revolution means there are potential endogeneity issues that authors elsewhere (Menaldo 2012) have tried to correct for using instrumental approaches). 

The DDR in particular is of interest since this was the partition of a country which had existed as a single entity pre-war. Englehart et al (2002), for instance, explicitly identify the division between the BRD and the DDR within a pseudo experimental context

The intergenerational transmission of the risk of divorce is a well-known long-term effect of divorce that has been found in many Western societies. Less known is the extent to which different family policies and divorce laws have an effect on the intergenerational transmission of divorce. In this paper, the division of Germany into two separate states from 1949 until 1990, with the consequent development of two very different family policies, is regarded as a natural experiment that enables us to investigate the effect of family policy on the mechanisms underlying the social inheritance of divorce [emphasis mine]. Data from respondents from the former East and West Germany participating in the German Life History Study are analyzed, using multivariate event-history methods. The results indicate that the strength of the intergenerational divorce transmission, when adjusted for differences in the divorce level, was lower in the East than in the West. Differences in marriage age and the timing of first birth, which are partial indicators of family policy, as well as differences in religion, could explain this effect. Furthermore, we found a tendency towards a reduction in the dynamics of divorce transmission over time, both in East and West Germany.

Engelhardt et al 2002. 

A contrasting but related argument is made by Konietzka and Kreyenfeld (2002), who argue that post Wende the replacement of the institutions of the DDR should lead to convergence of East German demographic behaviours with those in the West: the implicit assumption here it seems that there is a causal and exogenous effect of institutions of demographic behaviour 

In contrast to West Germany, where marriage and childbirth have been strongly coupled, the German Democratic Republic (GDR) displayed high rates of non-marital childbearing. Researchers attributed this pattern to “misguided” GDR family policies that encouraged women to remain unmarried after childbirth. With German unification, East Germany’s legal and political institutions — including family policies — were replaced by those of West Germany. Against this background, it was widely expected that east German non-marital birth rates would soon fall to west German levels.

Konietzka D, Kreyenfeld M. 2002. Women’s employment and non-marital

childbearing: a comparison between East and West Germany in the 1990s.

Population 57(2) 331-357

Socialism: an exogenous social institution?

So how good is this research design? According to Klüsener and Goldstein 2012, not very. There are two fundamental problems with the sort of pseudo-experimental design we were driving at earlier.
Firstly, to quote directly from  Klüsener and Goldstein:

The demarcation line between East and West Germany was an outcome of World War II. However, it is important to note that the border between the two German states did not emerge as an incidental freeze-frame of Allied combat positions at the moment of German defeat in the spring of 1945. Indeed, it was a preconceived demarcation line that had been drawn in advance by the Allied victors. The border was based on preexisting administrative entities, such as former federal states and historic territories of principalities (see also Kettenacker, 1989). Thus, while the inner German border became a formal state border in 1949, its roots go much further back in time. Most of the border followed lines that predate the Weimar Republic and the German Empire of 1871, and can be traced back to the emergence of German territorial states in early modern times

Source: Klüsener and Goldstein 2012

As such, the division of the DDR and BRD follows existing divisions which had existed before unification: it is likely that there were some differences between these state in terms of demography, social and cultural, and legal institutions. Subsidiary principles meant that legislation around marriage and cohabitation were decided at a provincial level. Even within the Prussian state there was some degree of heterogeneity between East and West with a contrast between the Preußische Allgemeine Landrecht, and Code Civil in Prussian Rhineland.

At the very best then, our natural experiment has a confounding issue.

The second problem for our research design is that there is some degree of endogeneity between the sort of social institutions post war and demographic behaviour: East German social policy dissociated marriage and childbearing primarily because the conditions for such a change were already in place” (Salles, 2006). In comparison to our first figure, we can observe the nascent differences between the East and West at the formation of the German Empire, long before Socialism and the DDR.

Non marital births within the German Empire, 1878

Source: Klüsener and Goldstein 2012

Legend denotes share of births to unmarried mothers

As such then, the imposition of Socialism and its subsequent collapse within Germany does not represent the natural experiment that we might have hoped for, were we trying to make statements about causal and exogenous effects of policy on demography. We need to look at little harder to find our identification strategy. 

Contraceptive choice and spousal communication

This post is based off a presentation I'm giving at LSHTM later on today, but previews a later presentation at PAA 2017. I've also used preliminary results in interviews, although the presentation version sadly lacks in both throwing tennis balls around and Star Trek references.

The premise we have here is that certain types of fertility awareness contraceptives have been associated with improved couple communication regarding family planning. A lot of this evidence seems to be based on qualitative evidence from India, where male partners reported greater willingness to become involved in family planning, or at the very least an interest in involvement in the regulation of fertility using this method. The exact logic model is not specified here: potentially we are either raising raising awareness regarding family planning or merely flagging opportunities for coitus, but in any case this does seem to present a beneficial side effect: the use of traditional contraception seems to be associated with improved couple communication which has all sorts of benefits.

The problem however, is that it's a reasonably easy to imagine non-causal explanations for this association: couples with better underlying communication are more likely to select methods which require higher levels of communication in the first place. This is particularly true of traditional contraceptives- it is much harder to practice covert contraceptive use using a traditional method and under circumstances where the woman has limited negotiating power almost impossible. We therefore get a classic selection type relationship resulting in an endogenous relationship: both the level of family planning communication and method choice are both outcomes of underlying couple characteristics.

The solution we use here is to take an instrumental variable approach. The instrument we are using here is religion: religious affiliation is associated with method choice (certain religious sects prohibit the use of certain methods) but there is no fundamental reason to believe that spousal communication should be associated religious group. The exact model is represented below: note that the red arrow is the underlying correlation between the error term and the variable of interest, which produces a biased regression estimate, but there is no association between religion and the error (exclusion criterion). 

In terms of operationalisation then: spousal communication is a binary indicator of whether a woman has spoken with her spouse within the past year about family planning, and the 'treatment' here is the contraceptive method type (traditional, modern, none). We make a string of comparisons across countries: these are broadly in line with a DHS analytical paper on spousal communication. Abridged results are in the table below, with coefficients on a probit scale (we need an individual level error to be able to get a correlation matrix) so numbers above zero indicate a higher probability of spousal communication

Controls: Residence, Educational attainment, partner's educational attainment,

 age, partner is older, polygynous marriage, infecund

The results come in three stages: Model I looks at the simple uncontrolled regression, which seems to find consistent improvement in the levels of communication between spouses who use a traditional method compared to non-users (6/7 countries). This relationship is robust to the introduction of a series of controls (Model II: 5/7 countries positive and significant). However, these regression set ups are still not accounting for our endogeneity problem: the beta estimates presented are still biased due to a correlation with the error term in our regression. 

Model III uses an instrument and the significant effect of traditional methods across countries collapses. Compared to Model II, two countries lose significance entirely and two are 'weakened' to show significance at the 5% level only. Benin is the only country where traditional methods improve spousal communication significantly at the 1% level. Interestingly, modern methods a robust to both the introduction of controls and the instrument.

The overall finding then is, I suppose, slightly disappointing: a purported advantage of traditional methods does not in fact exist and is largely explained by selection effects. I'm not sure that there are major policy implications here: spousal communication has never been a major reason for contraceptive method choice, and the purported benefit really was only a serendipitous side effect. For couples where the most suitable method is a traditional one, I'm not sure that this analysis should change that decision. Moreover, there are naturally limitations to this analysis- the most glaring being that the instrument we have is somewhat weak (measures of association are in the order of 0.20). However, it does fundamentally highlight the fact that selection effects exist within family planning and that failing to account for these can lead policy makers to overplay their hand when advocating certain solutions. If we want to use genuine evidence based policy we need to think about the nature of behaviours we are generating evidence about: simple regression may not be adequate. 

Research publication: Latent class models for cross-national comparisons

Latent class models for cross-national comparisons: The association between individual & national-level characteristics in fertility & partnership was published today in the International Journal of Population Studies. 

Abstract: Multilevel modelling techniques such as random models or fixed effect are increasingly used in social sciences and demography to both account for clustering within higher level aggregations and evaluate the interaction between individual and contextual information. While this is justifiable in some studies, the extension of multilevel models to national level analysis — and particularly cross-national comparative analysis — is problematic and can hamper the understanding of the interplay between individual and country level characteristics. This paper proposes an alternative approach, which allocates countries to classes based on economic, labour market and policy characteristics. Classes influence the profiles of three key demographic behaviours at a sub-national level: marriage, cohabitation and first birth timing. Woman level data are drawn from a subset of the Harmonized Histories dataset, and national level information from the GGP contextual database. In this example, three country classes are extracted reflecting two Western patterns and an Eastern pattern, divided approximately along the Hajnal line. While Western countries tend to exhibit higher levels of family allowances albeit accounting for a lower share of spending which is associated with lower marriage and later fertility, Eastern countries generally show a higher share of spending but at lower absolute levels with lower cohabitation rates and early fertility

Link to the full paper can be found here. This is open access, but I've included some thoughts below as well.

This paper is somewhat related to previous grumbles about the way in which demographic behaviour and institutions are modelled, or at least talked about. For full disclosure, this paper does not deal with the sort of endogeneity problems that I have talked about before: it does however I think provide a more coherant framework for analysing the interaction between national level policy and individual level demographic behaviour.

One of the major innovations in statistics has been the use of multilevel models. Random effects type models will typically deal with some degree of dependence of individual level outcomes on the context within which they are located. Examples of this can be found in schooling: there is variation in pupil level outcomes which is not explained by standard controls such as receipt of Free School Meals (poverty measure), SEN, ethnicity, etc.. Multilevel models are able to partition into individual (attributable to the pupil) and school level variation (This, as I understand it, is the basis of the Contextual Value Added measure in UK school league tables: the CVA is the school level residual in a random effects multilevel model). This is incorporated as an additional error term (often with a Normal distribution for convenience). While this works well for schools (there are enough of them that the school level errors can approximate a Normal distribution), whether this applies to countries is more questionable. It takes a lot of work to get even a dozen or so countries in a form where they can be modelled together (for instance see the excellent Harmonized Histories); but we are still below the sort of numbers where we can reasonably expect the CLT to apply. More problematic is the consistency between the model assumptions and the application: purposive selection of countries for analysis seems strongly at odds that our observations follow the sort of stochastic assumptions required for the model, the endogeneity problems seem to contradict iid assumptions of the residuals. In sum, the application of random effects models seems like too serious a violation of Gauss-Markov assumptions to be tolerable: obviously all models are wrong but there needs to be some sort of consistency between our theory and estimated models. We're not Milton Friedman. The other problem is that since we often want to make statements about cross national variation, the fact that this variation of interest is now subsumed into an error term tends to mean that actually interpreting variation between countries is tough

Fixed effects models are slightly better in that they require fewer assumptions of that nature, but still don't get all of the to addressing the sort of problems we might have. Usual concerns about statistical efficiency are probably not relevant here: the small number of countries probably means that there are not many efficiency gains to be made including a variance term over 7 dummies. However, is the fact that if we want to make statements about the effect of national level policy, the policy indicator is now fully confounded by the country fixed effect, so we have an identification problem. Unless we are including the fixed effect just to remove noise (so why are we taking a multilevel approach anyway?) this is a major research drawback.

The proposed solution is something of a hybrid: I use national level policy indicators to form latent classes (shown below) and then model demographic behaviours dependent on the country level class. 

Lyons-Amos (2016)

There are a number of advantages to this approach: we aren't purely tied to the reasonably strict assumption of the random effects model: the latent classes hare discrete to we can account for clumpiness in the 'residual.' Also, these classes are qualitatively interpretable: the use of policy indicators means that we can see what the class actually means. 

Table 3: Estimated mean levels of country level indicators by latent class 

Indicator (Number of country members)	Class 1: Eastern Europe (4)	Class 2: Western Europe lesser support (2)	Class : Western Europe higher support (4)
Family support
Value of family allowance (PPP adjusted 2005 $)	82.21	92.26	133.0
Family allowance (% of GDP)	1.38	0.11	1.78
Social expenditure (% of GDP)	13.16	26.8	26.13
Public expenditure on childcare (% of GDP)	0.38	0.64	0.38
Ease of childcare
Female labour force participation (%)	64.40	73.40	70.46
School entry age	3.25	4.00	3.00
Legal status of cohabitation
Cohabitation mentioned (%)	26.7	26.0	29.7
Legal equivalence (%)	30.7	37.5	34.5
Legally recognised (prob)	0.00	0.50	0.99

                                                                                                                                               

Lyons-Amos (2016)

Naturally there are some limitations in what we are able to do with these models, which I talk about in the paper: this is means to be a step on the way, not a complete solution. Whether this application is relevant as well remains to be seen- the added difficulty in estimating the latent class model will be a turn off for many. However, models need to be as simple as possible, but no simpler. I'm not sure whether for many demographers simply running MLwiN or xtmixed adequately addressed the latter concern. 

Inclusive institutions are endogenous institutions

This is post is a somewhat fleshed out version of a slightly snarky tweet based on an article published in Demographic Research. Before we proceed, I should say that this is not intended as a criticism of the article directly: the article is an example of somewhat wider trend in the demographic literature.

Equality at home - A question of career? Housework, norms, and policies in a European comparative perspective

Abstract

Background: Dual-earner families are widespread in contemporary Europe, yet the division of housework is highly gendered, with women still bearing the lion’s share. However, women in dual-career couples and in other types of non-traditional couples, across and within different European countries, appear to handle the division of housework differently.

Objective: The objective of this study is to examine the division of housework among various couple-earner types, by determining i) whether relative resources, time spent on paid work, gender attitudes, and family structure reduce variations in housework between different couple types, and ii) whether the division of housework varies between countries with different work‒family policies and gender norms.

Methods: The study uses data from ten countries, representing different welfare regime types, extracted from the European Social Survey (2010/11), and employs multivariate regressions and aggregated analysis of the association between the division of housework and the contextual indices.

Results: The results show that dual-career couples divide housework more equally than dual-earner couples, relating more to the fact that the former group of women do less housework in general, rather than that men are doing more. The cross-national analysis shows tangible differences between dual-earner and dual-career couples; however, the difference is less marked with respect to the division of housework in countries with more institutional support for work‒family reconciliation and less traditional gender norms.

Contribution: By combining conventional economic and gender-based approaches with an institutional framework, this study contributes to the research field by showing that the division of housework within different couple-earner types is contextually embedded

In and of itself, I'm rather happy with this abstract. I think there is a strong contribution here: the interaction between social institutions and demographic behaviour is clearly important and certainly underexplored. What I have a quibble with is the final modelling set up: the variable of interest here is the division of housework with a gender norm and work-family policy index as independent variables (plus controls). I am totally unconvinced by the idea that work-family policy is a control variable here: to me both household work and and work-family policy are endogenous outcomes of gender norms.

Example: Divorce reform in the United Kingdom
To take a further example, we'll examine marital dissolution in the United Kingdom. As a brief introduction, prior to 1969 in the United Kingdom, there was a rather onerous burden if couples wanted to divorce: proof was required of grounds for divorce, including adultery, drunkenness, insanity and cruelty. The 1969 divorce reform act changed this: couples could now divorce after 2 years separation with no 'fault.' Here's what happened to divorce rates:

Source: Office for National Statistics:Divorces in England and Wales: 2014 

Fair enough: here was can see that there is a (sharp) uptick in the number of divorces following the Divorce Reform Act: a reform to a legal institution resulted in a change in demographic behaviour.

The problem here is the extent to which we regard the change in the divorce law as exogenous to the actual behaviours that were occurring in the UK. We should note that the Royal Commission on Marriage and Divorce (1951-55) recommended against any reform of divorce law, arguing from moral principles that family life should be protected

The large number of marriages which each year are ending in the divorce court is a matter of grave concern...This disturbing situation is attributable to a variety of factors ...In the first place, marriages today are at risk to a greater extent than formerly. The complexity of modern life multiplies the potential causes of disagreement and the possibilities of friction between husband and wife...It must also be recognised that greater demands are now made of marriage, consequent on the spread of education, higher standards of living and the social and economic emancipation of women...Old restraints, such as social penalties on sexual relations outside marriage, have been weakened...[and there is] a tendency to regard the assertion of one's own individuality as a right, and to pursue one's personal satisfaction, reckless of the consequences to others...There is a tendency to take the duties and responsibilities of marriage less seriously than formerly

Royal Commission on Marriage and Divorce, 1956: 7-8 

We can see here however, recognition that behaviours were already changing: from our figure above the number of divorces is already increasing before divorce reform. Given then that laws are formed partially to reflect the cultural mores of the societies that they are governing (they are partially inclusive, rather than absolutist as per the Commission's recommendations) the Divorce Reform Act is an outcome of social shift (by the late 60s even the Church had recognised that fault was no a basis on which to grant secular divorce). This is also reflected in demographic behaviours- the legal institution and the behaviour it governs are endogenous. To that extent, this implies that any demographic article which is modelling policy and associated demographic behaviours needs to recognise this endogeneity: legal and policy institutions are as much an outcome variable as divorce rates. Let's model them that way. 

How bad demography gave me period pains

Recently, neurobiologist/personality scientist Adam Perkins published a book entitled The Welfare Trait where he claimed that employment resistant personality types are encouraged to breed by the current structure of the (UK) welfare state. Unsurprisingly, this claim was somewhat controversial, with criticism of a somewhat technical nature coming some media and some puff pieces of how the piece was being censored by 'the establishment' from the literal son of a baron.

Perkins' theory

The effects of this [1999 UK increase in child benefit] change to UK welfare provisioning have been studied in detail by Brewer et al. (2011) revealing that reproduction is more sensitive to changes in welfare legislation than had ever previously been shown: not only did these increases in the generosity of per-child welfare in the UK in 1999 increase the number of children born to benefit recipients by approximately 15% [emphasis mine] but also this effect was nuanced according to the specific opportunity-costs circumstances of the individual women.

The Welfare Trait, Chapter 4

As the quote above highlights, Perkins argues essentially that the effect of welfare is to increase the number of children born to families with employment resistant personality types, thus selecting children into this environment and increasing the prevalence of this trait within the population. Consider the following Lexis diagram which takes a hypothetical woman across her reproductive life course, with filled dots representing childbearing. In the absence of policy, she has two children as shown below. 

The paper by Brewer does indeed find that there is an increase in childbearing in response the introduction of more generous child benefits where childbearing is operationalised as having had a child within the previous twelve months. This leads Perkins to make the claim that the increased welfare does something like this, and we increase the number of children across the lifetime due to the additional birth (empty dot).

Why is this bad demography?

Perkins' mistake here is that he is using period (i.e. point or interval in time) evidence to make claims about cohort (or lifetime) effects. It should be stressed that there is a lot of evidence that increased welfare provision, particularly that targeted at children, does increase annual/period childbearing rates (e.g. Parr and Guest 2011). Additionally, there is some evidence that the economically disadvantaged (this is what Perkins means by "specific opportunity-costs circumstances of the individual women") are likely to make greater family orientated transitions (particularly females), that this is relatively independent of their economic outlook (as per Perkins' idea of employment resistance, or at least ambivalence) and that their fertility behaviour tends to depend on the welfare regime in which they are living. 

However, this is not sufficient evidence that this will cause the total number of children that a woman bears across her lifetime to increase: it is perfectly possible that a woman will bring forward her planned childbearing (grey arrow in the figure above) to take of advantage of welfare payments which may later be withdrawn, and have the same number of children across her lifetime as the counterfactual of no welfare reform. Indeed, most studies that look at completed childbearing tend to find that the effects are strongest in terms of timing (Bjorklund 2006) leading to a compression of birth intervals rather than increasing parity progression, with the effect of cash policies on long run childbearing being rather small (Gauthier and Hatzius 1996). Indeed, one of the only papers I am aware of that claims anything like a substantial effect in cohort fertility refers to Glastnost era Soviet Russia. I am unsure that this is a solid basis for making inferences to the 21st century UK.

Do teaching personalities vary? Reflecting the difference in student learning types onto teaching styles (Reblog from LSHTM PGCILT)

I recently delivered some taught materials as part of an ongoing favour. I’m not going to disclose the institution or course for privacy reasons, but for some context: I helped deliver this course while I was a postdoc at this institution, and have written and delivered materials for this course. On this occasion however, I was delivering a lecture and workshop on a slightly different topic and was sent some materials that someone else had prepared. Now, these materials I have to say were great- they took a reasonably technical topic and presented it clearly so people working at a variety of levels could access the materials. As well as the technical lectures there was an applied session where the students could implement the models being described on a real life dataset to address research questions. In terms of synthesising learning styles, this was top notch.

The problem however was that I found this material exceptionally difficult to deliver. Now, as a reasonably diligent demonstrator I had of course been through the materials beforehand, and knew what I was covering. I will again stress that I found the materials excellent- from the perspective from which I was receiving them for the first time: effectively as a student.

However, the experiences of the student and the teacher are fundamentally different. So while I had a decent understanding of the student experience, I didn’t really have an overview of the teacher’s perspective. I also fully take on board that there was something to take on board in terms of my teaching delivery- I’m rather more linear in terms of my delivery, which these materials eschewed in favour of breaking up the theoretical and practical elements so there is something to be learned here- but the fundamental point I feel is that in terms of teaching we need to understand this: not only are there different learning styles that we need to take on board when designing materials but there are differing teaching style as well, which we should accommodate for both others who might use our materials (think TAs here) but also for our own purposes.