1Be it births, deaths or marriages, most demographic phenomena have seasonal patterns of variation. While rarely studied from sociological viewpoint, these patterns are of enormous interest for understanding a society, its organization and its rites. “The timing of marriages reflects the rhythms of our collective life, and their transformations” (Besnard, 1989). This is perfectly illustrated by Jean Bourgeois’ study of marriage seasonality, published 70 years ago. His paper provides a valuable record of the society of his time, [1] a point to which we will return shortly. But this immersion in the past, and the historical depth it offers also arouse our curiosity. How has the distribution of marriages changed in the intervening years? What does it say about marriage trends, and the very institution of marriage itself? How would this study of seasonality be approached today?

2The marriage referred to in Jean Bourgeois’ article is of quite another time, when marriage was such an evident choice that the possibility of an alternative was never considered. Admittedly, between the industrial revolution which saw certain fringes of the working classes prefer consensual unions, and the decline of marriage first observed at the end of the post-war boom years, the period studied by Bourgeois (1927-1938) was a golden age for the institution of marriage. The religious dimension of marriage is even more striking. Readers today might get the impression that only religious marriage existed, Christian marriage to be more specific, the disconnection between the civil and religious components of marriage being absent from the analysis. [2] The variability of the religious dimension is addressed only indirectly, via the study of regional disparities. To account for these disparities, Jean Bourgeois surmises, for example, that “observance of Lent does not have, or perhaps no longer has, the religious importance that we tend to attribute to it” (p. 689). Religious precepts are referred to as “legislation”: Lent, the months of the Virgin Mary (May and August) and Advent, periods when weddings are, in principle, prohibited, largely shaped the seasonal pattern of marriages at that time. Economic factors also seem to be at play: the corn and grape harvesting months were not favourable for marriage. In this respect, Bourgeois’ article also tells us about the context of France, a country that was much more rural and agricultural than today.

3These various factors advanced by Bourgeois to explain marriage seasonality are equally valid for explaining the seasonality of births (Dupâquier, 1976; Houdaille, 1985): conceptions were fewer during religious festivals and in times of penitence, and likewise in periods of intense farm labour. The correlation between seasonality of marriages and births is thus explained by factors common to both, but it may also be the consequence of a dual effect. In the absence of effective contraception, the timing of marriage – which preceded cohabitation and first sexual intercourse at that time – could influence the seasonality of first births (de Saboulin, 1978); conversely, in a context where extra-marital births were condemned, a pregnancy could precipitate marriage (Lutinier, 1987). Bourgeois does not mention this phenomenon, perhaps because such social facts were still taboo in his day. This might also explain why the religious rules he mentions were often ignored; marriage during closed periods was actually quite frequent.

4The study of seasonality is always a good entry point for describing, and above all understanding, demographic phenomena and their patterns. Analysis of the current seasonality of marriages shows that seasonal variations are still large, but that the pattern is radically different today. Alongside a steep decline in the annual number of marriages (from 365,000 on average over the period 1946-1953 to 248,000 for the period 2006-2013), which clearly illustrates the changing place of marriage in society, the annual distribution of weddings reveals a different set of changes (Figure 1). First, there is a very clear weakening of the religious institution and of its authority over marital behaviour. The April peak and the May trough highlighted by Jean Bourgeois, reflecting observance of Lent, remain very pronounced up to the early 1950s, but grow smaller across successive marriage cohorts until they finally disappear. The November dip disappears likewise. A majority of marriages registered between 2006 and 2013, took place in the summer months, with six in 10 occurring between June and September, versus four in 10 over the period 1946-1953. In parallel with the lesser observance of religious precepts, church weddings now represent a small minority of total marriage numbers (70,000 in 2012, [3] i.e. fewer than three in 10).

Figure 1

Trends in monthly distribution of marriages, 1946-2013 (monthly index base 100)

Trends in monthly distribution of marriages, 1946-2013 (monthly index base 100)

Interpretation: An index of 120 (resp. 80) corresponds to a month where 20% more (resp. fewer) marriages were registered than in an average month (base 100).
Coverage: Metropolitan France.

5At a finer level of detail, if we examine the distribution of marriages day by day, apart from the regular peaks corresponding to weddings on Saturdays, a few “accidents” are visible, indicative of the importance that certain couples attach to the institution of marriage. Take the year 2012, for example (Figure 2). In February, we see a peak in weddings on Tuesday 14^th, St Valentine’s day, with 10 times more marriages than on the previous Tuesday. Another curiosity is visible on Wednesday 12 December, with 11 times more marriages than the previous Wednesday (5 December), because that date was the 12^th day of the 12^th month of 2012! While these examples are anecdotic, they illustrate the trend towards greater individualization that has characterized family transformations in recent years. People no longer marry in response to external norms dictating their behaviour, but are free to choose their spouse, the date of their wedding and the way it is celebrated. Coupledom and marriage are still the dominant mode of organization of private life, however, and social patterns in marriage remain strong. Its seasonality is a perfect illustration of this.

Figure 2

Daily number of marriages in February and December 2012

Daily number of marriages in February and December 2012

Note: The horizontal line indicates the daily average number of weddings in 2012 (655).
Coverage: Metropolitan France.

6Religious factors no longer explain the seasonality of marriages. Months that in the past were “ill-fated”, such as May, when “only donkeys marry”, according to an old saying cited by Bourgeois, and all the months of spring and summer more generally, are now the most popular wedding months. A contemporary sociological approach to marriage seasonality would not need to focus on the cult of Mary, Lent or the farming cycle. Other factors would now be considered. The importance of a “successful marriage”, for example, suggests a need to look for correlations between climate and marriage seasonality, since today “marriage seasonality … is more clearly linked to the solar calendar and, in part, to the school year” (Maillochon, 2016).

7The diversity of contemporary marriage practices would also need to be taken more fully into account. For Bourgeois, who worked with vital statistics, such an approach would be impossible. Today, large-scale sociological surveys reveal the multiple forms of marriage and enable us to explore, for example, the link between the lavishness of wedding celebrations and their timing. Likewise, its meaning for the marriage partners, ranging from a symbolic investment expressed via a large celebration and a simple administrative and legal contract, are likely to be correlated with its seasonality. Factors such as these provide valuable information for the study of contemporary marriage. The short history of civil partnerships (pacte civil de solidarité or PACS), first introduced in France in 1999, provides a good example of the importance of an approach based on seasonality. The tax rules governing PACS partners were changed twice (in 2005 and 2011), each time resulting in a marked change in the timing of PACS unions (Mazuy et al., 2016). More than a simple indicator of the observance of religious norms, the seasonality of marriage today reflects its heterogeneity and, more broadly, its social signification.