Divorce trends and variations along the rural-urban gradient in France, 1884-1913.

HAL Id: halshs-03480177

https://shs.hal.science/halshs-03480177

Submitted on 14 Dec 2021

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of sci-

entic research documents, whether they are pub-

lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diusion de documents

scientiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Divorce trends and variations along the rural-urban

gradient in France, 1884-1913.

Sandra Brée

To cite this version:

Sandra Brée. Divorce trends and variations along the rural-urban gradient in France, 1884-1913..

Revue Quetelet Journal, 2021, 8, pp.103 - 138. �10.14428/rqj2020.08.01.04�. �halshs-03480177�



Vol.8,n°1,2020,pp.103‐138

DOI:10.14428/rqj2020.08.01.04

ISSN:2593‐9157



Divorcetrendsandvariationsalongtherural‐urban

gradientinFrance,1884‐1913



SandraBrée



ThisworkislicensedunderaCreativeCommonsAttribution‐NonCommercial4.0InternationalLicense.Youcan

share,adaptthematerialfornon‐commercialpurposesprovidedthatyougiveappropriatecreditandindicateif

changesweremade.Fordetailsseehttps://creativecommons.org/licenses/by‐sa/4.0/







REVUE QUETELET/QUETELET JOURNAL

Vol. 8, n° 1, 2020, pp. 103-138

DOI : 10.14428/rqj2020.08.01.04

Divorcetrendsandvariations

alongtherural‐urbangradient

inFrance,1884‐1913

SANDRABRÉE



Résumé

Cetarticleseproposederelirel’histoiredesdivorcesdepuislerétablissementdu

divorceenFranceen1884jusqu’àlaveilledelaPremièreGuerremondialeendis‐

tinguanttroisgrandsensemblesdepopulation:lapopulationrurale,lapopulation

urbaineetcelledu départementdelaSeineconsidérés séparément.Pouraffiner

l’analyse,desdonnéesserontajoutéesquidistinguent,auseindudépartementde

laSeine,Parisdesabanlieue.Lessourcesdisponiblesfournissentparailleurs des

détailssurlesdivorces,engénéralindisponiblesautrementqu’àl’échellenationale,

tellesquel’épouxayantobtenuledivorce,lemotifdecelui‐ci,laduréedumariage,

l’âgeetl’écartd’âgedesépouxouencorelenombred’enfantsdescouplesdivorcés.

Plusieursrésultatsressortentdel’analyse.Lespopulationsurbainesdivorcentplus

quelespopulationsrurales,enparticulierdansledépartementdelaSeineetiln’exis‐

tepasd’homogénéisationdescomportementsentre1884et1913.Lescaractéris‐

tiques des divorces sont également différentes dans les territoires et distinguent

souventledépartementdelaSeinedurestedelapopulation.Enfin,undesapports

majeursdecetarticlepermisgrâceauxdonnéesdistinguantParisdesabanlieueau

seindudépartementdelaSeine,estdemettreenlumièrelesniveauxextrêmement

élevésdedivorcedanslabanlieueparisienne.

Mots‐clés

Divorce,France,populationurbaineetrurale,Paris,banlieue.

Abstract

Thispaperoffersarereadingofthehistoryofdivorcefromthereinstatementofdi‐

vorceinFrancein1884untiltheeveoftheFirstWorldWar,bycomparingthree

majorpopulationgroupings: the country’sruralpopulation,itsurbanpopulation,

andthatoftheSeinedepartment,consideredseparately.Torefinetheanalysis,data

willbeaddedthatdistinguishes,withintheSeinedepartment,Parisfromitssuburbs.

The available sources also provide detail on divorce that is generally unavailable

other than at the national level, such as which spouse obtainedthedivorce,the

1. Laboratoire de Recherche historique Rhône-Alpes (Lahra), Lyon, France.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

104

groundscitedforthedivorce,thedurationofthemarriage,theagesofandagedif‐

ferencebetweenthetwospouses,andthenumberofchildrenthemarriagepro‐

duced.

Severalresultsemergefromtheanalysis.Urbanpopulationsdivorcedmorethanru‐

ralpopulations,particularlyintheSeinedepartment,andtherewasnohomogeni‐

sationofbehavioursbetween1884and1913.Thecharacteristicsofdivorcealsodif‐

feredbetweenthepopulationgroupings,withtheSeinedepartmentoftendistinct

fromtherestofthepopulation.Finally,oneofthemajorcontributionsofthisarticle,

thankstothedatadifferentiatingParisfromitssuburbswithintheSeinedepartment,

istohighlighttheextremelyhighlevelsofdivorceintheParisiansuburbs.

Keywords

Divorce,France,urbanandruralpopulation,Paris,suburbs.

Introduction

In France, research on pre-Second World War separation and divorce

from a sociodemographic perspective is scarce, especially at the individ-

ual level. While aggregated studies taking into account spatial aspects at

the departmental (département) level do exist

, they often highlight the

urban manifestation of divorce: there is no research, to our knowledge,

that distinguishes between rural and urban municipalities (communes) to

complement the standard departmental breakdown. The Annual Statis‐

ticson PopulationMovement

distinguishes between three mutually ex-

clusive population groupings: France’s rural population, France’s urban

population, and the population of the Seine department (encompassing

Paris and its suburbs).

Divorce had been allowed in France between 1792 and 1816 and was

banned again until the 1884 Naquet Act reinstated it. Our analysis will

focus on the period from 1884 to 1913, after which data according to

these three groupings are no longer available. Our analysis will thus be

focused on the divorces in the initial period after the Naquet Act.

2. Thus, in terms of the period preceding the 1975 divorce reform law, Schnapper

(1978) spatially analysed legal separations for 1837-1884, Ledermann (1948) legal sep-

arations and divorces for 1936-1937, Roussel (1970) those for 1964-1967, and Munoz-

Perez (1981) those for 1970-1975. More recently, Brée (forthcoming) completed these

analyses for the period 1884-1936.

3. Statistique annuelle du mouvement de la population.

Sandra Brée

105

Apart from measuring the intensity of divorce among these three group-

ings, this analysis examines whether there was a convergence of divorc-

ing behaviour between 1884 and 1913. The data available for these

groupings – which is aggregate data – also provide details on divorce that

are generally only available in aggregate at the national level, such as

which spouse filed for the divorce, the grounds given for the divorce, the

duration of the marriage, the ages of the spouses and their age difference,

and the number of children they had. Another aim will therefore be to

find out whether the characteristics of divorcees are the same among

them and, if not, to endeavour to understand why they diverge.

These data represent the sum of the data at the municipal level, now lost.

They thus have the disadvantage of being highly aggregated, thus only al-

lowing for descriptive analyses. However, they do make it possible to pre-

cisely differentiate between the behaviours of these three populations,

rather than working according to a departmental breakdown, where

some departments are considered urban but also include rural areas and

suburbs and thus often behave very differently from central urban areas.

The analysis of the data across these three categories – the rural popula-

tion, the urban population, and the Seine department considered as a sep-

arate entity – even though they are aggregated, may be one of the only

ways to distinguish clearly between urban and rural behaviour, as this is

not possible through analysis by department what often contain both ru-

ral and urban populations – which differ greatly in terms of demographic

behaviour.

In addition, in order to further this analysis, the data available in the Sta‐

tisticalYearbooksoftheCityofParis

will be also pulled into the research.

These data, specific to the Seine department, distinguish Paris proper

from its suburbs within the department. This will provide additional nu-

ance to the distinction between the three basic groupings offered by the

AnnualStatisticsonPopulationMovement. Research on Paris’s suburbs is

rare even though it offers a valuable framework for analysis (Boudjaaba,

De Luca-Barrusse, 2013), especially from the second half of the 19th cen-

tury, when their populations grew explosively.



4. Annuaire statistique de la ville de Paris (since 1942 : Annuaire statistique de la

ville de Paris et des communes suburbaines de la Seine).

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

106

Literatureandhypotheses

It is highly likely that divorce was more common in urban than in rural

areas, and that divorce rates in the Seine department were particularly

high in comparison with the rest of the country. Phillips (1979) explains

a greater tendency towards divorce in urban areas during the periods of

the Revolution and the Empire as due to an increasing ease in finding

housing and employment (particularly with the dissociation between the

city of residence and the city of work), but also the rise in urban individ-

ualism and the weakening authority of the father and the family. Con-

versely, the more rural departments – where religious practice was

stronger and where a certain «traditional way of life [linked] to a complex

system of values, attitudes and relationships» was maintained (Roussel,

1970, p. 296) – seem to have been more resistant to marriage dissolution

(i.e., divorce or legal separation). It should be noted that in Flanders, men

– the effect does not appear for women – born in rural areas were found

to be less likely to divorce than those born in urban areas, leading the

authors of the study to posit that «rural socialization acts as a brake on

divorce» (Matthijs etal., 2008, p. 251). Brée and Gourdon (2020), how-

ever, rather show mobility as having a significant impact on divorce in

two suburban Paris communes: migrants born in the countryside were

more likely to divorce than those born in the communes under study. If

rural socialization may act as a brake on divorce, rural people who moved

to the cities seem to have had a relatively higher probability of divorcing,

perhaps because of greater detachment from their families and commu-

nities.

The link between socioeconomic level and risk of divorce is not always

obvious in the literature. In Flanders (Matthijs et al., 2008) and in the

Netherlands (van Poppel, 1997) in the 19th century, men working in an

unskilled or low-skilled profession seem to have been less likely to di-

vorce, partly due to their difficult socioeconomic situation: by staying to-

gether, the spouses could survive more easily with their two combined

incomes. The work of Kalmijn, Vanassche and Matthijs (2011) on divorce

in Flanders and the Netherlands between the late 19th and early 20th

centuries suggests that it is those with higher educational attainment, as

opposed to the more well-off, that were more inclined to divorce. In

France, it would seem that the middle classes have been most prone to

divorce, whether during the revolutionary period and the First Empire

(1792-1816) (Dessertine, 1981) or in the 1970s (Boigeol, Commaille,

1974). Between 1884 and 1933, on a national scale and from a quantita-

tive standpoint (and with the biases linked to statistical categorization),

Sandra Brée

107

it appears that there were many workers among the divorcees (Brée,

forthcoming). The wealthiest couples did not tend to divorce, nor did

farmers (Brée, forthcoming; Ronsin, 1990); among these groups it was no

doubt considered preferable not to split assets, especially if the couple

had children to whom they could pass on their property.

The effect of age of spouses and the age gap at marriage is broadly docu-

mented in the academic literature. The literature seems to show that mar-

riage at a young age, especially for women (Dessertine, 1981; Matthijs et

al., 2008; van Poppel, 1997), may have been an important factor in the

probability of divorce, but also the age difference between spouses, espe-

cially if the woman was older than the man.

The effect of prior marital status also plays an important role. In 19th-

century Flanders, wives who had already been divorced or widowed

were more than ten times more likely to divorce than the first-time mar-

ried (Matthijs etal., 2008). In The Hague, again in the 19th century, the

risk of divorce was four times higher if at least one of the partners had

already divorced once (van Poppel, 1997). In France at the beginning of

the 20th century, having already been divorced was a (low) risk factor for

getting divorced again, as in Belgium and the Netherlands; on the other

hand, unlike in Flanders, widowhood protected against divorce, for men

and even more so for women (Brée, forthcoming). For Matthijs, Baerts,

and van de Putte (2008), these effects could be linked to a large age dif-

ference, the presence of children, material interests, and potential con-

flicts. Brée and Gourdon (2020) add that widows and divorced women,

having already experienced the end of a union and more or less forced

celibacy, would know more about how to deal with such a situation than

first-time-married women, or at least be less fearful of the risks. Moreo-

ver, divorced women would by definition have knowledge of the legal

workings of divorce proceedings. Both factors, especially the latter, may

have alleviated fears about the prospect of a relationship breaking down,

potentially leading to a greater likelihood of divorcing.

The literature also shows that in France, as elsewhere, women were the

most frequent divorce applicants (Bertillon, 1883; Brée, forthcoming).

According to Bertillon (1883, p. 120), women asked for divorce because

men were «more often unbearable husbands than [the women] are some-

times unbearable wives». The differences in the gendered distribution of

divorce filings should instead be interpreted in terms of the benefits of

divorcing (despite gender inequality in terms of marriage as a system and

the economic difficulties that a divorce can generate; see Brée forthcom-

ing) and particularly the possibility of women to ensure their safety that

of their children. The stories they told in the divorce procedures (Sohn,

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

108

1996) show that a number of them were beaten, injured, or assaulted by

their husbands. Women sought protection in asking for divorce, but those

women who could afford to divorce without threatening their reputation

and economic status were also those with the education and legal and in-

stitutional knowledge to do so (Simonsson, Sandström, 2011). This can

refer to the willingness componentof Coale’s theory when applied to di-

vorce (Simonsson, Sandström, 2011). Coale’s theory (1973), summarized

by the formula «ready-willing-able», suggests that the adoption of a new

demographic behaviour is the result of a set of factors and requires three

prerequisites: individuals must consider the adoption of the new behav-

iour to be advantageous (they must be ready), it must be socially and mor-

ally accepted (they must be willing), and individuals must be in the posi-

tion to adopt it (they must be able). Simonsson and Sandström (2011)

applied this theory to divorce (but only on a national scale) by showing

that changes in the economic, social, cultural, legal, and political context

in Sweden had an impact on divorce rates.

The above-mentioned studies are based on statistical data that provide

information at the national level (and sometimes the regional level, but

only for divorce rates) or on individual data that provide very valuable

information but only for geographically or historically limited popula-

tions. However, there has been no research on the characteristics of di-

vorce among urban populations as opposed to rural.

Our research questions thus focus on different aspects of divorce in terms

of the groupings we are analysing:

– What are the divorce rates in the groupings studied, and do the gaps between

them narrow over the period 1884-1913?

– Is the proportion of women filing for divorce the same in all the groupings,

and how can this be explained?

– What grounds for divorce are used among the different groupings, and what

can they reveal about the different marital and family relationships that

characterize these groupings?

– How long do marriages last, and what does this say about the acceptance of

divorce?

– What are the characteristics of the divorcees in terms of age, age gap, previ-

ous marital status, family size, and socioeconomic profile?

– What are the specificities of the groupings (education, religion, migration,

age, socioeconomic level, illegitimacy) that could help explain the incidence

of divorce, the proportion of applicants that are women, and the grounds

given for divorce?

– In sum, what can a study differentiating between urban and rural popula-

tions contribute to the discussion on the socioeconomic aspects of divorce?

Sandra Brée

109

Sourcesandmeasures

Definitionsandsources

Until the end of the 18th century, the designation of the different types of

inhabited places was mainly based on the distinction between towns, vil-

lages, and hamlets, but with no precise definitions applicable at the na-

tional level. The term «rural» was not yet in use; the world of small towns

and villages dominated by farming families was referred to as la cam‐

pagne or the countryside (Bontron, 2015). Shortly after the Revolution

(1793), with the establishment of the communesystem, all town and vil-

lage names were abolished by decree and replaced by the name of the

commune in which they were located. It was only in 1846 that the Bureau

de la statistique générale de France began to distinguish between rural

and urban communes in the five-yearly censuses (Bontron, 2015). The

distinction was made as follows:

– To qualify as urban, a commune must have at least 2’000 inhabitants living

in its administrative centre.

– Communes are considered rural if fewer than 2’000 inhabitants live in the

administrative centre.

– The Seine department, encompassing Paris and its suburbs, is considered

separately, as a category of its own: all its communes are considered urban,

even those whose administrative centres have fewer than 2,000 inhabitants.

The distinction between France’s rural population, its urban population,

and the Seine department was made in the AnnualStatisticsonPopulation

Movement from its first edition in 1851 to 1913. The data produced by

this categorization cannot be replicated since this would require access

to the data at the commune level, which was discarded and is no longer

available. After the First World War, a distinction is only made between

the Seine department and the rest of France. The exception is between

1946 and 1952, when the classification was re-taken up and refined, with

the urban communes grouped by number of inhabitants: 2’000-5’000,

5’001-10’000, 10’001-50’000, 50’000-100’000, and more than 100’000.

However, the Seine department was then no longer considered in distinc-

tion to the other urban communes. Moreover, the data under this classi-

fication is less detailed, with no information on number of divorces, mar-

ital status prior to marriage, number of children at the time of marriage,

or indications on marriage duration. For these reasons, this article will

focus on the 1884-1913 period.

The AnnualStatisticsonPopulationMovement provide extensive informa-

tion on births, marriages, deaths, and divorces (from 1885 for the latter).

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

110

The data in relation to divorce do not come from the statistics on the court

rulings themselves (available in the GeneralAccountsofCivilJustice)

but

are based on transcriptions of divorce certificates in the civil status reg-

isters of the town halls where couples married (on average six months

after the divorce was pronounced). The following analysis will therefore

focus on the divorces of couples according to their place of marriage – and

thus capture the impact of the social environment at the time of mar-

riage – and not their place of divorce. According to Ledermann (1948),

these data are statistically reliable. However, a portion of divorces were

not recorded (10%) in the civil status registers because of the death of

one of the spouses or through negligence, and thus are not reflected in

theAnnualStatisticsonPopulationMovement (Ledermann, 1948).

Data from the AnnualStatisticsonPopulationMovementprovide, for the

three population groupings – France’s populations categorized as urban,

rural, or within the Seine department – from 1885 to 1913, data on the

number of divorces per month, the age of the spouses at the time of di-

vorce, the duration of marriages, and, for the first few years, the divorce

figures according to the occupations of the divorced men (Table 1). Two

volumes are particularly detailed (1907-1910 and 1911-1913) since they

provide, in addition to data on marital status prior to marriage, data on

the grounds on which divorce was granted, the spouse who obtained the

divorce, the number of children born of the marriage, and the number of

divorces dissolved for which a prenuptial agreement had been drawn up

or which had been preceded by legal separation; all these data are pro-

vided in relation to the age of the women.

Not all the data that appear in these specific volumes are available in the

other volumes of the publication. This means that most of the detailed

information about divorces are only available for the 1907-1913 period.



5. There are two main sources that can be used for the quantitative analysis of di-

vorce: the General Accounts of Civil Justice and the Annual Statistics on Population Move-

ment. Only the latter distinguishes between the groupings we are dealing with here. The

Annual Statistics on Population Movement summarize the transcriptions of divorce cer-

tificates in the civil status registers of the town halls where couples married, whereas the

data from the General Accounts of Civil Justice pertain to the requests (made and granted)

made in the courts where couples resided at the time of applying for divorce (Brée, 2019,

fothcoming).

Sandra Brée

111

TABLE1 Dataavailableinthesourcesforthegeographical

populationgroupings(1885‐1913)

Typeofdata

Populationgroupings

Urban Rural

Seinede‐

partment

ParisCity

Source

nnualStatisticson

PopulationMovement**

StatisticalYearbooks

oftheCityofParis***

Numberofdivorcespermonth 1885‐1913 1885‐1913 1885‐1913 1885‐1913

Ageofspouses* 1885‐1913* 1885‐1913* 1885‐1913* 1885‐1913

Agedifferencebetween

spouses

   1885‐1913

Durationofthemarriage 1885‐1913 1885‐1913 1885‐1913 1885‐1913

Husband’soccupation 1885‐1913 1885‐1913 1885‐1913 1885‐1913

Previousmaritalstatus 1907‐1913 1907‐1913 1907‐1913 1885‐1913

Groundsfordivorce 1907‐1913 1907‐1913 1907‐1913 1885‐1913

Spousewhoobtaineddivorce 1907‐1913 1907‐1913 1907‐1913 1885‐1913

Numberofchildren 1907‐1913 1907‐1913 1907‐1913 1885‐1913

Legitimacyofchildren    1885‐1913

Prenuptialagreement 1907‐1913 1907‐1913 1907‐1913 

Divorcesprecededbylegal

separation

1907‐1913 1907‐1913 1907‐1913 1885‐1896

Judgementinthefirstinstance

oronappeal

   1885‐1896

* IntheAnnualStatisticsonPopulationMovement,thedataforageofspouseiscross‐tabulated.

** IntheAnnualStatisticsonPopulationMovement,fortheperiod1907‐1913,alldataarebrokendown

byageofwife.

***IntheStatisticalYearbooksoftheCityofParis,thedataisavailableforthecityproperonly(withthe

exceptionofnumberofdivorces,availableforthelargestsuburbancommunesfrom1893).Inthese

publications,thedurationofmarriageiscross‐referencedwiththenumberofsurvivingchildrenfrom

theunionandtheoccupationofthehusbandwiththesexofthespousewhoobtainedthedivorce.

Also,thenumberofsurvivingchildrenfromthemarriageisnoted,butnothingisspecifiedintheAnnual

StatisticsonPopulationMovement.

The data for Paris, available in the StatisticalYearbooksoftheCityofParis,

are for the city proper only (with the exception of the number of divorces,

available for the largest suburban communes from 1893

), which makes

it possible to differentiate the city from its suburbs by subtraction. These

data provide information on the age of the spouses at the time of the di-

vorce, their age difference, marriage duration, the number of children

born of the union (and their legitimacy), the occupation of the husband,

the civil status of the spouses prior to the marriage, the grounds for di-

vorce, the sex of the spouse who obtained the divorce, and finally whether

the judgment was pronounced in the first instance or on appeal and

whether the divorce was preceded by legal separation (Table 1).

6. From 1893, a new section was added to the Statistical Yearbooks of the City of

Paris, entitled «Demographic Bulletin of the Communes of the Seine Department», provid-

ing information about the number of births, deaths, marriages, divorces, etc.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

112

The yearbooks provide data on Paris’s suburbs within the Seine depart-

ment from 1893 onwards (not shown in Table 1), but only on the number

of divorces (detailed for certain cities). The other information detailed in

Table 1 – age of spouses, previous marital status, etc. – is only available

in the yearbooks for the population of Paris city proper. We have there-

fore estimated the data for Paris’s suburbs from the two indirect sources

of information available to us (by subtracting the data for Paris from the

data for the Seine department), not a straightforward task.

Measures

Usually, specific divorce rates

are calculated by relating the number of

divorces over one year to the number of marriages over the same year.

This measurement is easy to make since the researcher only needs to find

the data for one year.

Another measurement

would consist in relating the number of divorces

in year Y to the number of marriages having taken place x years before, x

being the average duration of marriages that ended in year Y (12 years

on average for our study; see below). This measurement is complicated

by the fact that marriages do not last the same number of years every-

where and for all periods, and this is probably why it is not used very of-

ten. Another reason is that it is necessary to take into account circumstan-

tial events that took place in each of the two years in question, that of the

marriage and that of the divorce, to explain variations in divorce rates.

However, as Figure 1 shows, the comparison between the two indicators

shows substantial differences, especially for the urban populations, and

within the Seine department above all. This seems logical since the urban

areas, and especially greater Paris, gained many inhabitants between

1880 and 1913. Consequently, the number of marriages is higher in 1910

than in 1898 for instance. Therefore, the divorce rate calculated for the

same year is much higher than the one measured by relating the number

of divorces to the number of marriages 12 years before. Overall, the

movements are the same, but the gaps are stronger with the -12-years

indicator than with the one-year indicator. This should be remembered

7. The crude divorce rate is the number of divorces occurring among the popula-

tion of a given geographical area during a given year per 1’000 for the mid-year total pop-

ulation of that geographical area for that same year.

8. One can also measure the divorce rate relating the number of divorces per

10’000 married men or women, but we do not have the required data for the population

groupings we are working on here.

Sandra Brée

113

when analysing the data with the indicator we have chosen, the one with

a 12-year lag (see Figure 1).

IGURE

1 Divorceratesaccordingtotwodifferentmeasurements

Source:AnnualStatisticsonPopulationMovement,1885‐1913.

Note: The‐12curvesreflecttheratethatrelatesthenumberofdivorcesinyearyrelativetothenumber

ofmarriagesinyeary‐12.

Overviewofthegeographicalpopulationgroupings

France’s massive industrialization in the 19th century led to significant

urbanization. Consequently, the proportion of France’s population cate-

gorized as rural, which by 1861 had already declined to 71% (compared

with 82% in 1806; Dupâquier, 1989; Le Mée, 1989), fell to 55% in 1913.

At the same time, the proportion in urban areas increased: from 29% of

the population in 1851 (of which 5% were in Seine department) to 45%

in 1913 (of which 11% were in Seine). Within the latter, the population

of Paris proper accounted for 4.7% of the population of France in 1861

and 7% in 1913, and that of Paris’s suburbs 0.7% in 1861 and 3.4% in

1913.

Theagestructure is known to have been younger in the cities (especially

in the largest ones) than in the countryside, with the age structure of the

urban population showing an imbalance: adults aged 20-39 are overrep-

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

114

resented, and young people under 20 and adults over 60 are underrepre-

sented (Bourdelais, 1988; Bourillon, 1992). This is due to significantmi‐

gration from the countryside (and abroad) to the cities. In 1891, 55.6% of

the suburban population was born outside the Seine department and

«barely one person in five is of commune origin» (Farcy, 1991, p. 41). This

proportion is much higher than that of the capital, which stood at 36.4%

in 1891 (39% in 1911). According to some estimates, in 1914, 8% of the

urban population was of foreign origin and 75% of rural origin (Garden,

Le Bras, 1988). Nevertheless, it is mainly Paris and its metropolitan area

that carried weight in France’s urban population, with the other large cit-

ies remaining quite small in comparison because of much weaker

growth

. Indeed, if we take out the Paris agglomeration, 63% of the

French population was still rural in 1911 (Garden, 1988).

Despite their younger age structure, urban populations tended to experi-

ence excess mortality (Bocquier, Brée, 2018; Biraben, 1975; Dupâquier,

1990; Eggerickx, Debuisson, 1990), linked in particular to sanitation and

population density. In the rural population, while the high proportion of

young children and older people were major factors in the overall mor-

tality rate, mortality was still lower than in urban areas.

Astocrudebirthrates,they were higher in provincial urban areas than in

rural areas prior to 1836 but did not differ much in urban and rural areas

thereafter, being slightly lower in urban areas from 1836 to 1872 and

slightly higher at the end of the century. During the first decade of the

20th century, the levels are the same in the rural and urban areas in Pro-

vincial France and even in the Seine department (respectively 20‰,

19.7‰, and 18.7‰ for the period 1904-1913). However, crude birth rate

are highly dependent on the structure of the population (Brée, 2017)

Tugault (1975) showed that the trends are the same for fertility, but the

reversal seems to have taken place a little after what is observed for the

birth rate. Fertility in provincial towns was, on average, slightly higher

than in the countryside around 1860, and it was during the second half of

the 19th century that the relatively large differences in fertility between

cities and the countryside widened, leading to urban underfertility and

rural overfertility, with the Seine always having much lower fertility lev-

els than everywhere else. Similarly, the crudemarriagerate was higher in

9. About the urban transition in provincial France, see Bocquier and Bree (2018).

10. Brée (2017) shows that the birth rate in Paris was higher than for all of France

during the 19th century, while its fertility rate – which takes into consideration the age

structure of the population – was much lower.

Sandra Brée

115

the Seine department – and particularly in Paris – than elsewhere in

France

. However, when the number of marriages is compared to the so-

called «marriageable» population, i.e., the number of adults aged 15-60,

the marriage rate was actually lower in the Seine department than else-

where in France.

TABLE2 Proportionofmarriagesper100marriageable

peopleintheSeinedepartmentandFrance’s

otherdepartements(singleadultsaged15‐60)

 Seine Otherdepartments France

1874‐1878 6.28 7.32 7.22

1879‐1883 5.86 6.66 6.59

1884‐1888 6.04 6.32 6.30

1889‐1893 5.98 6.50 6.44

1894‐1898 6.12 6.50 6.46

Source:AnnualStatisticsonPopulationMovement,1899.

The marriage rate is lower in the cities, but non-marital cohabitation

(concubinage) is more frequent in the largest ones, such as Paris (Battagl-

iola, 1995), and also in the most industrialized ones, such as Le Creusot

(Bourdelais, Demonet, 1998). Due to concubinage but also the high num-

bers of what was termed «filles‐mères» (single mothers) whose children

were not recognized by the father – nonmarital childbearing was much

more frequent in the cities than in the rural areas. Out-of-wedlock births

accounted for 23.2% of births in the Seine department between 1904 and

1913, 12.2% of births in urban areas, and 4.6% of births in rural areas. In

other words, illegitimacy was twice as high in the Seine department as in

other urban areas, and six times as high as in rural areas. Within this de-

partment, illegitimacy was slightly higher in Paris (26% in 1902) than in

the suburbs (20% in 1902) (Brée, 2014, 2017).

As to socioeconomicstructure, unsurprisingly the rural population mainly

worked in agriculture: in 1911, 68% of rural households were considered

as farming households (Molinier, 1977) and consisted of around 35% of

the working population (Marchand, Thélot, 1991). In the cities, 45% of

the working population worked in non-agricultural sectors (Marchand,

Thélot, 1991). Cities had many more workers but also people working in

liberal or administrative occupations. In the suburbs, the working class

11. It was 93‰ for the Seine department between 1875 and 1913 against 73‰ for

France’s urban population and 72‰ for its rural population. Unfortunately, the data does

not distinguish between the urban and rural populations.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

116

was also very numerous. In the case of the Seine department, at the end

of the 19th century, 10% of the active population worked in agriculture

– and this population was relegated to the outer reaches of the depart-

ment, far from Paris – 44% worked in industry and 21% in trade, and

12% lived off of annuities (Brée, 2015).

The levelofeducation differed widely across the three groupings. People

in urban areas, and even more so in the Seine department (especially

Paris), were much better educated than people living in the countryside,

even though literacy inequalities diminished, and almost disappeared,

between 1869 and 1906 (Table 3).

TABLE3 Proportionofwomenandmensigningtheirmarriage

certificatein1869,1885,and1906intheSeinedepartment

(Parisandsuburbs)andamongtherestofthecountry’s

ruralandurbanpopulations

Seine Paris Urbanpop. Ruralpop.

Men

1869 96.0 97.2 79.8 71.6

1885 98.3 99.7 90.5 84.4

1906 100 100 98.0 97.0

Women

1869 88.4 91.6 66.2 59.8

1885 96.7 99.3 82.3 76.4

1906 99.0 99.9 96.0 95.0

Sources: AnnualStatisticsonPopulationMovement,1885(fortheyears1869and1885)

and1906;MunicipalStatisticsoftheCityofParis,1869;StatisticalYearbookof

theCityofParis,1885and1906.

Finally, it is known that the rural population was more attached to the

CatholicChurch than the urban population (Boulard, 1982; Morlet, 1990).

In 1911-1912, Lenten practices were still well entrenched in the country-

side (during the periods of Lent and Advent in the religious calendar dev-

otees abstained from sexual relations). Advent was no longer widely ob-

served, and there was even a peak in marriages in December in the Seine

department (Segalen, 1988). Indeed, religious practice, which was greatly

reduced in Paris in the 18th century (Chaunu etal., 1998), weakened even

further over the 19th century (Boudon, 2001). In Paris’s suburbs, Easter

practices were weaker than in Paris at the end of the century (Boulard,

1982).



Sandra Brée

117

DivorceinurbanandruralFrance

AshorthistoryofdivorceinFrance

Until the Law of Floréal year II (20 September 1792), which introduced

divorce by mutual consent, for incompatibility of temperament and other

«specified reasons» (raisonsdéterminées), only separation from bed and

board could break a marriage and remarrying was not allowed. The new

law was quickly criticized for being too liberal, and when the Civil Code

was adopted in 1804, the possibility of divorce was maintained but the

rules were tightened (Goy, 1988; Ronsin, 1990). The procedure was made

more strict and the formalities more rigorous, and as a result divorce be-

came exceptional from then on (allowing only for divorces due to «fault»).

Eventually divorce was abolished in April 1816, under pressure from the

Church. After many campaigns for reinstatement – notably in 1830, in

1848, and during the Paris Commune of 1871 (Ronsin, 1992) – it was fi-

nally achieved 27 July 1884 with the promulgation of the Naquet Law.

The legitimization of personal fulfilment, the right to love, and the right

for a woman’s claim to be heard were much less emphasized in the 1884

law than they were in the claims of the Divorciaires (proponents of legal-

ized divorce) earlier in the 19th century (Ronsin, 1992). This law upheld

the spirit of a sanction imposed on a spouse who has not stood by his or

her commitments and was guilty of a serious fault (adultery, criminal con-

viction) or of excessive, abusive, or insulting behaviour. It was not until

almost a century later, in 1975, that divorce by mutual consent was rein-

stated.

As soon as divorce was reauthorized applications came flooding in: 3’322

in 1884 alone, whereas the law was enacted in July. By way of compari-

son, legal separations amounted to about 3’000 per year in the early

1880s. In 1885, more than 4’000 divorces were granted; the figure was,

nearly 7’900 ten years later, 10’850 in 1904, and 15’113 in 1912 (Brée,

forthcoming).

Urbandivorce

The spatial analysis of divorce between 1884 and 1913 shows a predom-

inance of divorce in an area located in the Paris basin and to its north,

from Haute-Normandie to Champagne-Ardenne, as well as in the Medi-

terranean basin (Brée, forthcoming). Beyond this spatial distribution, the

most urbanized departments were Seine (Paris’s department), Rhône

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

118

(Lyon’s), Bouches-du-Rhône (Marseille’s), and Gironde (Bordeaux’s).

Conversely, some regions had very low divorce rates, such as Brittany,

the Basque country, and the southern Massif Central.

As in previous periods

, divorce seems to have been rather an urban phe-

nomenon; however, while the most urbanized departments had, on aver-

age, the highest divorce rates, there was no systematic correlation be-

tween level of urbanization and level of divorce (Brée, forthcoming;

Ronsin, 1990). Roussel (1970) points this out again for divorces granted

between 1936 and 1967.

Data show that between 1884 and 1913, 44% of divorces took place in an

urban commune, 24% in the Seine department, and 31% in a rural com-

mune, while the proportional distribution of marriages was 30%, 11%,

and 59%, respectively (and the populations 28%, 8%, and 55%, respec-

tively). Much higher divorce rates in urban areas, especially in the Seine

department, as already mentioned, were indeed emerging and deserve

further analysis.

For the period 1885-1900, the proportion of divorced people per 10’000

inhabitants was 10.3 in the Seine department, 4.4 in the urban popula-

tion, and 1.4 in the rural population. In 1885, the year following the rein-

statement of divorce, the proportion of divorced people was 15 times

higher in the Seine department than in the countryside and 6 times higher

than in the cities. These remarkably high rates are the result of the long

period during which couples could not divorce (1816-1883) and thereaf-

ter remarry. These ratios shrank to 5 and 2 times higher, respectively, by

1900.

Specific divorce rates (12-year lag) were twice as high in the Seine de-

partment as in the rest of France’s cities and seven times as high as in

rural areas (Figure 2). Between 1885 and 1913, the ratio between the di-

vorce rate in towns and in the Seine department was fairly stable, around

2, although it fell slightly (1.9 in 1886 compared with 1.7 in 1913). While

the population of the Seine had a divorce level twice as high as that of

France’s other cities, the divorce growth rate between 1886 and 1913

was fairly similar for both groups (1.9 for the Seine department, 2.2 for

urban communes). On the other hand, divorce in the countryside, which

12. During the Revolution and the Empire, most divorces took place in Paris (50 to

60 percent of all divorces) and in large cities, such as Rouen, Marseille, Lyon, Bordeaux,

and Toulouse, whereas they were very rare in smaller towns or in the countryside

(Ronsin, 1990). Paris remained predominant afterwards for legal separations, accounting

for 13.5% to 27% of separations between 1837 and 1883 (whereas its marriages ac-

counted for 3% to 7% of all marriages) (Brée, forthcoming).

Sandra Brée

119

was rare, grew at a slightly faster rate of 3.6. Despite this slightly higher

growth rate in rural areas, we cannot characterize the behaviours as ho-

mogenizing, and the curves remained parallel through the period.

Looking more specifically at the Seine department, it appears that, con-

trary to what might be expected, the divorce rate was equivalent in Paris

and its suburbs (the curve for suburban divorces was much more uneven

because the numbers were quite low): 7.8 divorces per 100 marriages

contracted 12 years earlier for the period 1885-1913.

IGURE

2 Divorceratesinthegeographicpopulationgroupings

between1885and1913

Reading:Numberofdivorcesinyearnrelativetothenumberofmarriagesinyearn‐12.

Sources: AnnualStatisticsonPopulationMovement,1885‐1913;StatisticalYearbookoftheCityofParis,

1885‐1913.

Whogetsdivorced,andonwhatgrounds?

Outside times of war, women are more likely than men to file for divorce

and to obtain it (Brée, forthcoming)

. For the period for which data are

13. The General Accounts of Justice provide data on the spouse who applied for the

divorce while the Annual Statistics on Population Movement data refer to the spouse(s)

who obtained it.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

120

available for the population segments under study (1907-1913), women

in France obtained and therefore presumably filed for; divorce (Brée

forthcoming) in 51.3% of cases, but they were slightly more likely to ob-

tain it in the Seine department (54%) (Table 4). The gaps between this

department and the rural and urban populations were, however, fairly

small. On the other hand, there seems to be a significant gap between the

Paris suburbs and the other segments, since women in the Seine depart-

ment obtained divorce much more often than elsewhere. As these figures

have been calculated based on statistical yearbooks for two different

years, however, they should be taken with a grain of salt.

TABLE4 SpousewhoobtainedthedivorceintheSeinedepartment

(Parisandsuburbs),France’sruralpopulation,itsurbanpopulation,

andforallofFrance(average1907‐1913)

Spousewhoob‐

taineddivorce

Seine

dept.

Seine:

Paris

Seine:

Suburbs

Urban

population

Rural

population

Allof

France

Man 39.7 41.2 34.3 42.6 42.2 41.8

Woman 54.1 53.0 58.1 49.9 51.2 51.3

Both 6.1 5.7 7.5 7.4 6.6 6.9

Total 100 100 100 100 100 100

Numberofdivorces 20’534 16’068 4’466 40’085 29’011 89’630

Undeclared 131 126 5 631 161 923

Source: AnnualStatisticsonPopulationMovement,1907‐1910and1911‐1913;StatisticalYearbooksof

theCityofParis,1907‐1913.

Returning to research hypotheses, the link between education level (Ta-

ble 3) and the proportion of women being the ones obtaining the divorce

(Table 4) is not truly clear or systematic. Because there is no reason to

suppose that women are more likely to be injured or assaulted by their

husbands in rural or urban areas (Tanguy, 2015), following (Sohn, 1996),

explains that this phenomenon does not seem to belong to a particular

environment or be more urban than rural

), this confirms the hypothesis

that divorce may be more accepted in some areas more than others.

The grounds for the divorces granted also vary across the three groupings

studied, with the Seine department clearly distinguished from the rest of

14. Note the link Bertillon (1883, pp. 71-72) makes between divorce, suicide, and

insanity. He explains that a country with a high degree of social malaise has lot of marital

discord and thus a higher suicide rate, and that ever-increasing deranged behaviour is

related to the stresses of modernity. He adds that the big cities, centres of human thought

and activity, produce more madmen, more drunks, and more suicide (and also more un-

bearable husbands) than do the countryside and small towns.

Sandra Brée

121

France (Figure 3). As Bertillon (1883) and Ronsin (1990) point out, it

seems that couples state the grounds that will allow them to divorce as

quickly and cheaply as possible, even if these grounds do not (or do not

completely) correspond to reality. However, the grounds for divorce in

the three jurisdictions studied are invoked with very different frequen-

cies. While «excesses, abuse, and serious insult» is the grounds most often

invoked throughout France, it is much more often given in the Seine de-

partment (72%) than elsewhere in France (46%-47% in both rural and

urban populations). Conversely, abandonment of the family home, for

both men and women – and this almost-even gender distribution is nota-

ble – is given as grounds four times more often outside the Seine depart-

ment, whether among rural or urban people (11%-12%), than in the de-

partment itself (3%). Finally, adultery by men is cited twice as often as

adultery by women in the rural and urban populations, whereas in the

Seine department, this ratio is much lower (1.4).

IGURE

3 GroundsfordivorcefortheSeinedepartment,otherurban,

andruralpopulationofFrance(average1907‐1913)

Source:AnnualStatisticsonPopulationMovement,1907‐1910,1911‐1913.

Unfortunately, the StatisticalYearbooksoftheCityofParis do not describe

the motives in as much detail as do the Annual Statistics on Population

Movement, but they do provide the official motives of divorces, broken

down as follows for the period 1907-1913: excesses and abuse (17.9%),

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

122

serious insult (62.1%), female adultery (8.4), male adultery (11.2%), and

criminal conviction (0.4%). It is therefore difficult to compare these data

with those available for the Seine department and the urban and rural

populations. It may simply be noted that the proportion of divorces

granted for female adultery and male adultery are very similar, which im-

plies that the same is true for the suburbs.

Marriageduration

Marriages that end in divorce were somewhat shorter (Figure 4) in the

Seine department (11 years on average between 1890 and 1913) than in

the rest of the urban (11.8 years) and especially rural (12.9 years) popu-

lation. Seine couples had a higher divorce rate after 5 to 10 years of mar-

riage (34% of marriages) than did rural couples (27%), who in turn had

a higher rate after 20 years of marriage (19% compared with 11% for

Seine couples).

IGURE

4 Meanmarriagedurationatthetimeofdivorce

forthegroupingsstudied

Reading: Numberofyearselapsedbetweentheyearofmarriageandtheyearinwhichthedivorceruling

istranscribedinthecivilregister.

Sources: AnnualStatisticsonPopulationMovement,1885‐1913;StatisticalYearbookoftheCityofParis,

1885‐1913.

Sandra Brée

123

Married couples in urban areas were three times more likely to divorce

with prior judicial separation than those in the Seine department, and

couples in rural areas were four times more likely (3% of divorces pre-

ceded by judicial separation in the Seine department, 9% in urban areas,

and 12% in rural areas in 1907-1913). The longer marriage durations of

urban and especially rural populations can therefore be explained in part

by a higher rate of transition to divorce via legal separation, whereas

Seine couples were more likely to turn to direct divorce. This may indi-

cate greater social acceptance of divorce in the Seine department than

elsewhere, whereas in rural areas, attempts were made to postpone di-

vorce as much as possible, particularly through legal separation, which is

less radical since the couple remained married and legal separation did

not allow for remarriage.

In the Seine department, suburban couples clearly divorced much earlier

than others (deduced from the fact that Paris city couples divorced later

than couples in the rest of the department

), whereas the marriage du-

ration in Paris city was closer to that of the rest of the urban population

(this is particularly true for the period 1889-1906). Research conducted

on two cities in Paris’s suburbs shows that marriages broken by divorce

between 1885 and 1912 in these two cities were indeed very short, 8.2

and 9.9 years, respectively (Brée, Gourdon, 2020). Couples in the Paris

suburbs were more likely to divorce and did so sooner than in the rest of

France. Brée and Gourdon (2020) show that those who moved to the cit-

ies they studied statistically divorced more than those born in them; they

argue that a sense of detachment from communities of origin could be an

important factor in the higher divorce rate and thus in their social ac-

ceptance. The proportion in the suburbs of those born elsewhere is

higher than it is in Paris and could indeed explain this difference in be-

haviour, which cannot be explained by education levels since Paris’s pop-

ulation has, on average, a better level of education than its suburban pop-

ulation.

Thus it seems that divorce was even better accepted in the suburbs than

in the capital. While this statement may come as a surprise, since large

cities are often seen as places where innovative behaviours are adopted

(birth restrictions, for example), it should not be forgotten that at the end

of the 19th century the proportion of children born out of wedlock was

15. When cross-checking the data published in the Annual Statistics on Population

Movement for the Seine department and those published in the Statistical Yearbooks of

the City of Paris for the city of Paris in order to derive data for the Paris suburbs alone,

the results sometimes seem inconsistent. We have therefore chosen to work only on the

data published in both publication series.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

124

particularly high in Paris’s suburbs and close to levels in Paris city (Brée,

2014). Also noteworthy is the higher proportion of children born to un-

married couples in the suburbs than in Paris (Brée, forthcoming). Sub-

urbs may thus be a place where more «marginal» behaviours such as hav-

ing children out of wedlock, concubinage, and divorce are more socially

accepted. Here we can probably distinguish innovative behaviours from

«marginal» ones that are not practiced by exactly the same population

even if the two types of behaviours have urban features.

Thecharacteristicsofdivorceinthepopulationgroupings

Ageofdivorcingspousesandagedifferences

Age at divorce is determined by the combination of age at marriage and

duration of marriage. Men and women divorced at younger ages in the

Seine department and among the urban population than among the rural

population but also Paris

(Figure 5).

For the urban population and the Seine department, the age at divorce

was about the same, although age at marriage

was a little older for the

Seine department and marriages lasted a bit longer among the rest of the

urban population. But what is most surprising is that age at divorce in

Paris and among the rural populations is the same. Age at first marriage

was higher in Paris than among the rural populations (an almost two-

year difference for men and more than two years for women), but mar-

riages lasted much longer in the rural areas.

16. Women’s age at marriage was around 25 years old for the Seine department, 23

years and 4 months to 23 years and 10 months for the urban population, and 23 years to

23 years and 6 months for the rural population. For men, ages at marriage were respec-

tively 28 years and 10 months to 29 years and 8 months for the Seine department, 27

years and 5 months to 28 years for the urban population, and 27 years and 8 months to

28 years for the rural population.

17. Women’s age at divorce was 33 years and 10 months between 1894 and 1913

in the Seine department, 34 years for urban populations, and 34.5 years for rural popula-

tions. For men, ages at divorce were, respectively, 38 years and 6 months for both the

Seine department and the urban populations and 39 years and 8 months for the rural

populations.

Sandra Brée

125

IGURE

5 Meanageofwomenatdivorceforthegroupingsstudied(1885‐1913)

Sources: AnnualStatisticsonPopulationMovement,1885‐1913;StatisticalYearbookoftheCityofParis,

1885‐1913.

IGURE

6 Meanageofmenatdivorceinthegroupingsstudied(1885‐1913)

Sources: AnnualStatisticsonPopulationMovement,1885‐1913;StatisticalYearbookoftheCityofParis,

1885‐1913.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

126

There are also age differences between spouses that differ across the pop-

ulations (Figure 7). For the urban population, including Paris and the

Seine department, for 80% of the marriages the husband is older and for

20% the wife is. On the other hand, for the rural population, as in subur-

ban Paris, men are more likely to be older than their wife. Obviously, no

conclusion can be drawn without some means of comparing differences

in age at marriage. Unfortunately, while the ages at marriage of the two

spouses are indicated in the Annual Statistics on Population Movement,

they are given for five-year age groups, so we cannot know the age differ-

ence between the spouses. The information is partially available for

1910-1913 since we know whether the husband was older than the wife

but not by how many years, and it is not available for 1895-1901, which

would allow us to compare couples divorced in 1907-1913 to their mar-

riage cohort. Nevertheless, if we compare these data with great caution,

it appears that among couples married in 1910, the husband was likely to

be older than among those divorcing in the same period for urban (77.9

versus 76.5%) and especially rural (85.2 versus 82%) populations. Con-

versely, they were a little likely to be older in the Seine department (75

versus 76.3%). Surprisingly, even if the husband is more often older in

the rural population, it is not there that the larger age gaps when the man

is older are observed but in suburban Paris and the urban population

(36%-37% of men older by 10 years or more), whereas it is in Paris that

women are more often the oldest (20%).

It is difficult to know whether this is actually linked to the fact that among

divorcing couples, men are slightly less likely to be older (which would,

however, be consistent with the idea that divorcing couples have smaller

age differences than others). However, these differences are small and

probably not very significant.

The StatisticalYearbooksoftheCityofParisprovide statistics on age dif-

ferences in marriage certificates. They show that among the divorcees of

1907-1913, men were slightly older than among all the couples who mar-

ried in 1895-1901, and women were likely to be older

. Couples seem to

be more likely to divorce when the man is older, and mainly when he is 5

to 9 years older (but not 10 years older or more), and less likely to divorce

when the woman is older. The gender imbalance favours divorce when

18. In Paris, 6% of couples married in 1895-1901 were the same age, compared with

5% of divorced Parisians in 1907-1913; for 30% and 31% of couples, respectively, men

were older by 1 to 4 years; 28% and 32% were older by 5 to 9 years, and 14% were older

by more than 10 years in both groups; the woman was 1 to 4 years older for 13.7% of

couples married in 1895-1901 in Paris and 11.8% of divorced Parisian couples in 1907-

1913 and, respectively, 7.4% and 5.8% older by 5 or more years .

Sandra Brée

127

the husband is older but prevents it when the wife is older. In contrast,

couples with a small age difference therefore were not more likely to di-

vorce than others and were even very slightly underrepresented.

IGURE

7 Agedifferenceofdivorcedspousesforthethreegroupingsstudied

(divorcedcouplesin1907‐1913)

Sources: AnnualStatisticsofPopulationMovement,1907‐1913;StatisticalYearbookoftheCityofParis,

1907‐1913.

Previousmaritalstatusandremarriageofdivorcedpeople

In all the groupings, people whose marriage was their first are overrepre-

sented among individuals whose marriages ended in divorce compared

to their marriage cohort (Table 5). However, the gap is much higher in

the Seine department than in the rest of the urban population and espe-

cially than in the rural population. Conversely, a far lower proportion of

widowers were divorced in the Seine department than elsewhere. As for

those previously divorced, their probability of divorcing again was lower

than others in the Seine department but higher in other cities and espe-

cially in the countryside.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

128

ABLE

5 Previousmaritalstatusofthosedivorcedbetween1907

and1913andthosemarried12yearsearlier(%)



Men Women

Seine

dept.

Paris

only

Urban

pop.

Rural

pop.

Seine

dept.

Paris

only

Urban

pop.

Rural

pop.

Marriedbe‐

tween1895

and1901

Single 85.6 87.8 88.9 92.0 87.4 89.5 91.4 94.9

Widowed 11.4 9.6 9.9 7.7 9.8 8.1 7.5 4.8

Divorced 2.9 2.6 1.2 0.3 2.8 2.4 1.1 0.2

Divorcedbe‐

tween1907

and1913

Single 93.6 93.4 92.2 93.6 94.4 94.0 93.5 94.9

Widowed 4.3 4.4 5.9 5.1 3.8 4.0 4.8 4.0

Divorced 2.1 2.2 1.9 1.2 1.8 2.1 1.7 1.1

Sources: AnnualStatisticsonPopulationMovement,1895‐1901&1911‐1913;StatisticalYearbookof

theCityofParis,1895‐1913.

Reading:IntheSeinedepartment,amongthemenmarriedbetween1895and1901,85.6%were

singlewhentheymarried,11.4%widowed,and2.9%divorced.Amongmendivorcedbe‐

tween1907and1913,inthesamedepartment,93.6%weresingleattheirmarriage(esti‐

mated12yearsearlier,i.e.,duringtheperiod1895‐1901),4.3% widowed,and2.1% di‐

vorced.

IGURE

8 Proportionofpreviouslydivorcedmenandwomen

amongthemarriedpopulation,1884‐1913

Source: AnnualStatisticsonPopulationMovement,1884‐1913;StatisticalYearbooksoftheCityofParis,

1884‐1913.

Sandra Brée

129

In sum, the fact of having previously divorced seems to have offered pro-

tection against a second divorce in the Seine department but to have fa-

voured a second divorce for the rest of the French population. Widow-

hood also protected against divorce, especially in the Seine department.

Divorced people did not remarry everywhere the same rate, but every-

where men remarried more than women (the differences were, however,

very small in Paris and the Seine department). Their share among mar-

ried people in the Seine department was much higher (and slightly higher

in the suburbs than in Paris) than among married people in other cities

and even higher than among married people in the countryside (Figure

8). Of course, this proportion depends, in particular, on the rate of remar-

riage of widowers for the different populations studied (but widowers

and widows were more numerous than elsewhere among the married, so

the effect is the opposite) but also on their age structures. Thus, the later

age at divorce in the countryside may explain why fewer people remar-

ried there than in the cities, especially in Paris, but it is also likely that

marriage to a divorced man or woman was better accepted in urban areas

than elsewhere. In view of the appreciable differences between each pop-

ulation, we can hypothesize that it was more conceivable, or even accept-

able, to marry a divorcee in the Seine department than in other urban ar-

eas and especially in the countryside. Moreover, the gender gap in couple

involving a divorcee remarrying was smaller in the Seine department

(even if men were more likely to remarry than women) than in urban and

especially rural areas, which supports the hypothesis of social accep-

tance.

Socioeconomiclevelofthosedivorced

TheAnnualStatisticsonPopulationMovementprovide, between 1885 and

1890, the distribution of divorces according to the husband’s occupation

(Appendix 1). This statistic is not really instructive unless it can be com-

pared to the distribution of occupations of married people in the same

cohorts. Unfortunately, this information is not given in the source.

It is possible to compare the proportion of previously divorced people

who signed a marriage prenuptial agreement with the proportion of

those that had done so 12 years earlier (average marriage duration) with

the limitation that these data are not available in the StatisticalYearbooks

oftheCityofParis for those divorced.

It appears that couples who divorced between 1907 and 1913 were

less

likely to have signed a prenuptial agreement at the time of their marriage

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

130

than all couples in their marriage cohort (Table 6). This result therefore

runs counter to our hypotheses, since it seems that it is not the wealthiest

households, nor those who could rely on their personal wealth, that were

most likely to divorce – quite the contrary. It is possible that this result is

linked to the fact that prenuptial agreements were more frequent among

rural populations, perhaps for couples with real estate assets that could

not easily be broken up. Nevertheless, if we look at the rural population,

we see that the gap between married and divorced couples for the pres-

ence of a prenuptial agreement is very small. Could the prenuptial agree-

ment therefore be an indicator of the proportion of the population that

held assets, whether agricultural or real estate? We would then return to

the conclusion that those who had personal wealth had a low probability

of divorce (Brée, forthcoming)

, but without any further implications ex-

cept that the gap between the share of married people having signed a

prenuptial agreement and the share of divorced people having done so

was much higher in the urban populations and the Seine department.

TABLE6 Proportionofcoupleshavingsignedaprenuptialagreement

(married1895‐1901;divorced1907‐1913)

 Seinedept. Urbanpopulation Ruralpopulation AllofFrance

Married1895‐1901 19.4 25.4 31.1 26.3

Divorced1907‐1913 14.6 18.2 30.2 21.3

Relativedeviation ‐25% ‐28% ‐3% ‐19%

Sources:AnnualStatisticsonPopulationMovement,1885‐1890.

FamilySize

Many divorced people did not have children at the time of their divorce

(Figure 9). This could be related to their age at divorce, where couples

had not necessarily had the time to have children. However, infertility

could also be a cause of divorce. Conversely, having one or more children

may have encouraged some couples not to divorce. Fertility differences

were significant among the three populations studied. It is in the Seine

department that the proportion of childless couples among the divorced

was highest, followed by the urban and then the rural population. Since

fertility in Paris was much lower than fertility in the cities, which in turn

had lower fertility than rural areas, it is hard to know whether the di-

vorced were broadly representative of the rest of their marriage cohort.

19. Like the better-off, they turned more to legal separation than to divorce (Brée,

forthcoming).

Sandra Brée

131

Since data are available by age group, however, we observe that three-

quarters of couples who divorced before the age of 20 did not have chil-

dren. Then in the next age group, the proportion of childless couples fell,

before rising again for divorces involving women between the ages of 40

and 49. Conversely, couples that divorced after the wife had passed her

childbearing years (about age 50) were mostly childless: 65%, or even

72% for those over 60 years of age. However, the proportion of these cou-

ples that divorced late in life without children was much higher in the

Seine department (87%) than in other urban areas (77%) and especially

in the countryside (59%).While it may seem that this could be linked to

the proportion of widows and widowers or divorced people in these

higher age groups who did not necessarily have children after remar-

riage, in fact widowers were more numerous in the rural population than

in the urban population or in Paris. It therefore seems that childlessness

favoured divorce, whether it was to eventually remarry and have the pos-

sibility of having children with the new spouse or the fact that having chil-

dren was an obstacle to divorce and that childless people could divorce

more easily. This childlessness of those divorced was much higher in the

Seine department.

IGURE

9 Numberofchildren(%)atthetimeofacouple’sdivorce,1907‐1913

Sources: AnnualStatisticsonPopulationMovement,1907‐1913;StatisticalYearbookoftheCityofParis,

1907‐1913.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

132

Conclusion

Populations residing in urban communes, and particularly in the Seine

department, began to divorce earlier and at a greater rate than their rural

peers. The divorce curves of these groupings were parallel from the end

of the 1880s and still were on the eve of the First World War.

Our data also show that the trend observed for the divorce rates is the

same as for the duration of marriages and the probability of remarrying:

all these indicators are higher in the Seine department than in the rest of

the urban population, and higher in the urban than in the rural popula-

tion. All this may point to a greater social acceptance of divorce in the

Seine department than elsewhere, whereas in rural areas, attempts were

made to postpone divorce as much as possible, particularly by means of

legal separation, which was less radical since the couple remained for-

mally married and were not permitted to remarry. However, there is very

little difference in the gender of the spouse who obtained the divorce,

even if the proportion of women who did so in the Seine department is a

little more than in urban and rural areas. Gender differences in receiving

the divorce are the same for urban and rural populations, and the distri-

bution of divorces by grounds is also almost equivalent in these two pop-

ulations. Everywhere the most common grounds cited is «excesses,

abuse, and serious insult», but much more so in the Seine department

than elsewhere in France. These differences are linked to the fact that

home abandonment was not often cited in the Seine department, unlike

in the other populations. In summary, there is a gradient according to the

urban hierarchy: the more urban the territory, the higher the divorce

rates and the shorter the duration of marriage. However, the Seine popu-

lation differs from the others in terms of low gender differentiation for

the filing of the divorce but especially of the grounds for divorce, whereas

for these last two elements, the urban and rural populations show similar

behaviour. The Seine therefore still stands out from the other two terri-

tories in terms of the specific grounds cited for divorce and, as expected,

adopted divorce faster and more emphatically than the others, with a

slightly higher involvement of women (but the difference with the other

populations is less than expected here). More surprisingly, some indica-

tors distinguish urban from rural populations (divorce rate, marriage du-

ration) and others do not (the gender of the spouse who obtained the di-

vorce and the grounds cited for divorce). It would be worthwhile doing

further research to try to understand why.

The distinction between Paris and its suburbs within the Seine depart-

ment nuances these findings. The level of divorce was equivalent in Paris

Sandra Brée

133

and in its suburbs, and most of the available data seem to indicate a higher

acceptance of divorce in the suburbs than in Paris (shorter duration of

marriages, greater likelihood of divorced men and women remarrying,

greater proportion of women obtaining the divorce). A major contribu-

tion of this paper is to bring to light the extremely high levels of divorce

in suburban Paris. This brought us to re-examine the question of the

adoption of innovative and marginal behaviours. City centres are often

seen as places where modern behaviours are first adopted, such as birth

control. They are also places where «marginal» behaviours are more

common, such as having children out of wedlock or concubinage, but

these behaviours are also found a lot, if not more, in the suburbs.

This leads to the question of the specificity of the geographical population

groupings studied in this paper that could explain the high, or low, levels

of divorce. We have noticed that the levels of divorce were higher where

education levels were highest. However, we can find no clear link be-

tween the proportion of women filing for divorce and their level of edu-

cation. As to the link between divorce and secularization, our research is

in line with the literature that shows that divorce rates are higher where

religion is little practiced. This distinguishes the rural and urban popula-

tion, but also the population within the Seine department, since seculari-

zation is even higher in suburban Paris than in Paris city. We have also

pointed out that the proportion of migrants is higher for the populations

where divorce is frequent, including within the Seine, where the propor-

tion of migrants in the suburbs is even higher than in Paris city proper.

Our research does not provide any significant results at the socioeco-

nomic level. We can only suggest that those populations among which di-

vorce is lowest are perhaps those where rates of transmission of inher-

itance are highest, which would tend to lead to an avoidance or a delay of

divorce (the rural population has a higher tendency to opt for legal sepa-

ration before proceeding to divorce).

All this broadly points in the same direction: divorce seems to be more

accepted in the urban communes, especially in Paris and maybe even

more in its suburbs. Detachment from the family and the community

brought about by migration but also detachment from religion may be

important factors in a higher acceptance of divorce (perhaps more than a

higher level of education). Also, migration may have another impact: mi-

grants may try to marry rapidly after migration or may have fewer op-

portunities in terms of partner choice, and this could also lead to more

divorces and shorter marriages. This is in line with previous findings at

the individual scale (Brée, Gourdon, 2020) and encourages further re-

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

134

search on the demographic behaviours of the urban and suburban popu-

lations, and especially on what could be termed «marginal» behaviours

such as out-of-wedlock childbearing, concubinage (and the separation of

these non-married couples), and divorce, especially since these behav-

iours would become more and more widespread, especially after the Sec-

ond World War and during the second demographic transition.

References

BATTAGLIOLA F. (1995), «Mariage, concubinage et relations entre les sexes. Paris,

1880‐1890»,Genèses.SciencesSocialesetHistoire,18(1),pp.68‐96,https://doi.org/

10.3406/genes.1995.1277.

ERTILLONJ.(1883),ÉtudeDémographiqueduDivorceetdeLaSéparationdeCorps

DanslesDifférentsPaysdel’Europe,Paris,Masson.

IRABENJ.‐N.(1975),«Quelquesaspectsdelamortalitéenmilieuurbain»,Population

3,pp.509‐522,https://doi.org/10.2307/1531350.

OCQUIERP.,BREES.(2018),«ARegionalPerspectiveontheEconomicDeterminants

of Urban Transition in 19th‐Century France», Demographic Research, 38 (50),

pp.1’535‐1’576,https://doi.org/10.4054/DemRes.2018.38.50.

OIGEOLA.,COMMAILLEJ.(1974),«Divorce,milieusocialetsituationdelafemme»,

ÉconomieetStatistique,53(1),pp.3‐21,https://doi.org/10.3406/estat.1974.1605.

ONTRONJ.‐C.(2015),«Ladimensionstatistiquedelaruralité»,Pour,228(4),pp.57‐

67,https://doi.org/10.3917/pour.228.0057.

OUDJAABAF.,DELUCA‐BARRUSSEV.(2013),«Populationsdebanlieue,unehistoireà

revisiter»,AnnalesdeDémographieHistorique,2013/2,pp.5‐15,https://doi.org/10.

3917/adh.126.0005.

OUDON J.‐O. (2001), Paris, Capitale religieuse sous le Second Empire, Paris, Cerf,

https://doi.org/10.4000/books.psorbonne.906.

OULARDF.(1982),Matériauxpourl’Histoirereligieusedupeuplefrançais,19ème‐

20èmesiècles,Paris,PressesdeSciencesPo.

OURDELAISP.(1988),«Levieillissementdelapopulation»,DUPAQUIERJ.(ed),Histoire

deLaPopulationFrançaise,TomeIII.De1789à1914,Paris,PUF,pp.230‐242.

OURDELAISP.,DEMONETP.(1998),«Lafréquenceduconcubinagedansunevillein‐

dustrielle:leCreusotaumilieudu19èmesiècle,questionsdeméthodesetrésul‐

tats»,B

RUNETG.,FAUVE‐CHAMOUXA.,ORISM.(eds),LeChoixDuConjoint,Programme

PluriannuelenSciencesHumainesdelarégionRhône‐Alpes,Lyon,pp.47‐57.

OURILLONF.(1992),LesvillesenFranceau19èmesiècle,Paris,Ophrys.

Sandra Brée

135

BREES.(2014),IncidencedelaféconditéillégitimesurlaféconditégénéraleàParis

au19èmesiècle,Espace,Population,Société,2014(1),https://doi.org/10.4000/eps.

5648.

REES.(2015),Lapopulationdelarégionparisienneau19èmesiècle,Démographie

etsociétés,DocumentdeTravail6,Centrederechercheendémographie,Louvain‐

la‐Neuve,49p.

REES.(2017),Parisl’inféconde :LalimitationdesnaissancesenRégionParisienne

au19èmesiècle,Paris,INEDÉditions,Études&EnquêtesHistoriques,https://doi.

org/10.4000/books.ined.1496.

REES.(2019),«Quellessourcespourl’analysedesdésunions(1830‐1970)?»,Paper

presentedattheConférenceUniversitaired’ÉtudedesPopulations,Dijon.

REES.(forthcoming),«DeuxsièclesdedésunionsenFrance(1792‐1975)»,Annales

deDémographieHistorique,2021(2).

REES.,GOURDONV.(2020),«Lespremiersdivorcésdelabanlieueparisiennesousla

TroisièmeRépublique.L’apportdel’analyseindividuelleàl’étudedudivorce»,Anna‐

lesdeDémographieHistorique,2020(2),https://doi.org/10.3917/adh.140.0129.

HAUNUP.,FOISILM.,DENOIREFONTAINEF.(1998),LeBasculementreligieuxdeParisau

18èmesiècle:Essaid’histoirepolitiqueetreligieuse,Paris,Fayard.

OALEA.J.(1973),«TheDemographicTransitionReconsidered»,InternationalPopu‐

lationConference,Vol.1,Liège,InternationalUnionfortheScientificStudyofPopu‐

lation,pp.53‐72.

ESSERTINE D. (1981), DivorceràLyonSousLaRévolutionetl’Empire, Lyon, PUL,

https://doi.org/10.4000/books.pul.13700.

UPAQUIER J.(1989),«LepleinruralenFrance»,Espace, Populations Sociétés, 3,

pp.349‐356,https://doi.org/10.3406/espos.1989.1341.

UPAQUIERJ.(1990),«Lasurmortalitéurbaine»,AnnalesdeDémographieHistorique

(Démographiedesvillesetdescampagnes),pp.7‐11,https://doi.org/10.3406/adh.

1990.1754.

GGERICKXT.,DEBUISSONM.(1990),«Lasurmortalitéurbaine:lecasdelaWallonieet

deBruxellesàlafindu19èmesiècle(1889‐1892)»,AnnalesdeDémographieHistori‐

que,(Démographiedesvillesetdescampagnes),pp.23‐41,https://doi.org/10.3406/

adh.1990.1756.

ARCYJ.‐C.(1991),«Banlieue1891:Lesenseignementsd’unrecensementexemplai‐

re»,F

AUREA.(ed),Lespremiersbanlieusards,Paris,Créaphis,pp.15‐69.

ARDENM.(1988),«Postface.LapopulationdelaFranceàlaveilledelaPremière

guerremondiale»,D

UPAQUIERJ.(ed),HistoiredelaPopulationFrançaise,Paris,PUF,

pp.503‐518,https://doi.org/10.3917/puf.dupaq.1988.01.0503.

ARDENM.,LEBRASH.(1988),«Ladynamiquerégionale»,DUPAQUIERJ.(ed),Histoire

deLaPopulationFrançaise.TomeIII.De1789à1914,Paris,PUF,pp.138‐161.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

136

GOYJ.(1988),«LaRévolutionfrançaiseetlafamille»,DUPAQUIERJ.(ed),HistoiredeLa

PopulationFrançaise,Paris,PUF,pp.84‐115.

ALMIJNM.,VANASSCHES.,MATTHIJSK.(2011),«DivorceandSocialClassduringthe

Early Stages of the Divorce Revolution: Evidence from Flanders and the Nether‐

lands», Journal of Family History, 36 (2), pp.159‐172, https://doi.org/10.1177/

0363199011398436.

EMEER.(1989),«LesvillesdeFranceetleurpopulationde1806à1851»,Annales

dedémographiehistorique,pp.321‐393.

EDERMANNS.(1948),«LesdivorcesetlesséparationsdecorpsenFrance»,Popula‐

tion,2,pp.313‐344,https://doi.org/10.2307/1523884.

ARCHAND O., THELOT C. (1991), Deux siècles de travail en France, INSEE Études,

https://doi.org/10.3406/estat.1990.5499.

ATTHIJSK.,BAERTSA.,VANDEPUTTEB.(2008),«DeterminantsofDivorcein19th‐Cen‐

tury Flanders», Journal of Family History, 33 (3), pp.239‐261, https://doi.org/10.

1177/0363199008319373.

OLINIER J.(1977),«L’évolutiondelapopulationagricoledu18èmesiècleànos

jours», Économie et Statistique, 91 (1), pp.79‐84, https://doi.org/10.3406/estat.

1977.3127.

ORLETJ.(1990),«Anciennesetnouvellespratiquesreligieusesdesruraux»,Archives

deSciencesSocialesdesReligions,72(1),pp.167‐185,https://doi.org/10.3406/assr.

1990.1357.

UNOZ‐PEREZB.(1981),«Larépartitiongéographiquedesdivorcesde1970à1975»,

AILLOND.,COSTECALDEN.,GODING.,MUNOZ‐PEREZB.(eds),LedivorceenFrance,Paris,

MinistèredelaJustice/INSEE,pp.99‐108.

HILLIPSR.G.(1979),«LedivorceenFranceàlafindu18ème



siècle»,Annales.Écono‐

mies, Sociétés, Civilisations, 2, pp.385‐398, https://doi.org/10.3406/ahess.1979.

294052.

ONSINF.(1990),Lecontratsentimental.Débatssurlemariage,l’amour,ledivorce,

del’AncienRégimeàLaRestauration,Paris,Aubier.

ONSINF.(1992),LesDivorciaires.Affrontementspolitiquesetconceptionsdumaria‐

gedanslaFrancedu19èmesiècle,Paris,Aubier.

OUSSELL.(1970),«LesdivorcesetlesséparationsdecorpsenFrance(1936‐1967)»,

Population,25(2),pp.275‐302,https://doi.org/10.2307/1529200.

CHNAPPERB.(1978),«Laséparationdecorpsde1837à1914.Essaidesociologiejuri‐

dique»,RevueHistorique,259,pp.453‐466.

EGALENM.(1988),«Lemariage»,DUPAQUIERJ.(ed),HistoiredeLaPopulationFrançai‐

se,TomeIII.De1789à1914,Paris,PUF,pp.423‐435.

Sandra Brée

137

SIMONSSONP.,SANDSTRÖMG.(2011),«Ready,Willing,andAbletoDivorce:AnEco‐

nomic and Cultural History of Divorce in Twentieth‐Century Sweden», Journal of

FamilyHistory,36(2),pp.210‐229,https://doi.org/10.1177/0363199010395853.

OHNA.‐M.(1996),Chrysalides.Femmesdanslavieprivée(19ème‐20èmesiècles),

Paris,PublicationsdelaSorbonne.

ANGUYJ.‐F.(2015),«LesvictimesdeviolencesconjugalesenBretagneau19èmesiè‐

cle»,G

ARNOTB.(ed),LesVictimes,desoubliéesdel’histoire?Rennes,PressesUniver‐

sitairesdeRennes,pp.259‐277,https://doi.org/10.4000/books.pur.18610.

UGAULTY.(1975),FéconditéetUrbanisation,Paris,INED,PUF,https://doi.org/10.

2307/1531033.

VAN POPPEL F. (1997), «Family Breakdown in 19th‐century Netherlands: Divorcing

CouplesinTheHague»,TheHistoryoftheFamily,2(1),pp.49‐72,https://doi.org/10.

1016/S1081‐602X(97)90010‐5.

Divorce trends and variations

along the rural-urban gradient in France, 1884-1913

138

Appendix1: Distributionofdivorcesfrom1885to1890

accordingtotheoccupationofthehusband

(%) Seinedept. Urbanpop. Ruralpop. AllofFrance

Militaryandmarine 0.6 2.4 0.8 1.5

Civilservants 3.7 4.0 3.9 3.9

Landlordsandrentiers 10.0 5.7 7.4 7.4

Liberalprofessions 7.1 5.2 3.2 5.3

Farmersandsettlers 0.8 2.5 14.1 5.0

Industrialbosses 9.3 8.3 6.9 8.2

Merchants 11.6 10.1 6.7 9.7

Commercialandindustrialemployees 18.8 15.2 7.3 14.2

Industrialworkersanddaylabourers 27.7 32.6 24.0 28.9

Farmworkersanddaylabourers 1.8 7.2 18.4 8.5

Domesticstaff 2.7 1.9 2.2 2.2

Other 5.9 4.9 4.9 5.2

Total 100.0 100.0 100.0 100.0

Source:AnnualStatisticsonPopulationMovement,1885‐1890.