1In the 1920s and 30s, one of the major threats looming over Europe was a low birth-rate. Throughout the old continent, the unrelenting decrease in the number of births was becoming cause for concern (Treves, 2001). In Italy, as in the rest of Western Europe, the decline in fertility had, by this time, expanded its scope from the small precursory groups to encompass the entire population (Livi Bacci, 1977 ; 1986). Governing bodies feared that this situation would soon result in population decline.

2In Italy, the causes of this phenomenon were initially attributed to the consequences of the First World War and the Spanish flu, which had claimed many victims of childbearing age (Pietra, 1929), but it soon became apparent that the real reasons were much deeper-rooted and longer-las-ting. [2] In the political and cultural context of the time, all across it was therefore felt necessary to implement measures to invert this tendency (Van Bavel, 2010).

3Fascist Italy was the country at the forefront of this quest. From the mid-1920s, the regime effectuated a series of interventions aimed at boosting the birth-rate, and contemporaneously launched a large-scale propaganda campaign for ruralisation, a topic which was complementary and closely linked to the ‘family policy’ (Ipsen, 1997 ; Treves, 2001). On the basis of the analyses carried out, necessarily at an aggregate level, these policies, at least in the initial phase, were deemed a failure by demographers of the time, and by Benito Mussolini himself.

4Successive demographic studies of this period are also often limited to the national, aggregate level, lacking the necessary detail to analyse differentials in fertility rates which are decisive in accounting for the variety of behaviours and reproductive models in Italy and that are often limited to the assertion and verification of the urban-rural dichotomy. For this reason, the regime’s pro-natalist interventions have continued to be judged negatively, while it is possible that they did achieve their desired effect in certain contexts or for certain socio-economic categories. Using individual data on four communities taken from the Census of 1961, this study aims to verify if the regime’s propaganda campaign and regulative measures to boost the birth-rate had any real impact within particular segments of the population.

5To achieve this objective, this paper is divided into four sections : the first retraces the main stages of the policies adopted by the regime to increase fertility levels ; the second examines the phases and characteristics of the decline in the birth-rate in Italy in order to contextualise the data of the four communities under study ; the third describes certain characteristics of the four communities during the fascist era ; and the fourth carries out a micro-level analysis, based on the reproductive behaviour of women from four communities, with the aim of highlighting the role of fascist population policies.

The policies of the Regime to increase the Birth-Rate

6The date that marks the start of the pro-natalist policy of the fascist regime in Italy is 26 May 1927, when Benito Mussolini gave his famous Discorso dell’Ascensione (“Ascension Day Speech”). On this occasion the Duce, addressing the House of Commons, dwelled upon the European and Italian decline in fertility, stigmatized industrial urbanisation, which he asserted “was making the population barren”, and claimed that “All the Nations and Empires had felt the bite of their decline, when they witnessed a fall in their number of births”. From that moment the guidelines of the fascist policy on the demography of the country were laid down. [3] The measures that followed were the introduction of a tax on bachelors, which came into effect on 1 October, 1927, and tax exemptions for large families, enacted in 1928.

7Great attention to the possible effects of these policies was immediately raised even outside Italy, in particular amongst biologists, social scientists and, above all, the many supporters of a “renewing” eugenics (Cassata, 2011). In 1928 the Italian Committee for the Study of Population Problems (CISP, Comitato italiano per lo studio dei problemi della popolazione) was founded, which promoted the need for population policies even at the international level. In the same year, also taking into account the negative findings brought to light by the “gross rate of reproduction” – the new measuring instrument of fertility codified by Kuczynski – Gini legitimised the natalist stance sanctioned by the Discorso dell’Ascensione (Gini, 1928). Two years later, in 1930, in an article that appeared in The Journal of Political Economy –, the real international scientific presentation of the fascist population policy –, Gini pointed out that “Italian population policy has its roots in recent scientific research in vital statistics and the evolution of nations”. [4] For Gini, many nations in Northern, Central and Western Europe displayed signs of biological decline and “in any event, it is vital for a nation like Italy, which finds itself on the margin of the zone of demographic depression, to avoid being drawn in, and to face the future in a condition of utmost demographic potentiality. Here is the scientific basis of the propulsive population-policy adopted in Italy by the will of Mussolini” (Gini, 1930, 690). In the closing paragraph, dedicated to an accurate description of the measures adopted to contrast the feared decline of the population, Gini himself appears to show certain cautiousness on their effectiveness. “The whole campaign against the fall in births was begun only recently, too recently, perhaps, to permit measurement of its results. Indeed, these results may never be estimated accurately, since the trend of the birth-rate and the rate of natural increase should, for this purpose, be compared with the hypothetical course which would have been if Mussolini’s propulsive policy had not been attempted. However, it can safely be stated that this policy has been beneficial, in as much as it serves as a sort of dike for Italy, protecting her from those neo-Malthusian tendencies which are undermining the future of other nations. If it meets with adequate response in the national consciousness, it will prove a great advantage to Italy in the contest of nations for survival” (Gini, 1930, 697).

8In the same year, a second series of events contributed to further delineate the Italian position regarding eugenics and population policy. In December, 1930, Pius XI radically condemned birth control, premarital certificates, abortion and sterilization with the encyclical Casti Connubii (Dau Novelli, 2003). In the following months, Agostino Gemelli, founder and rector of the Catholic University in Milan and vice-president of the SIGE (Società Italiana di Genetica ed Eugenetica), defended and actively promoted the argumentations of the papal encyclical. “Undoubtedly, the institutional, ideological and political agreement between the fascist regime and Catholic Church – sanctioned in 1929 by the Lateran Treaty – was decisive in the affirmation, in Italy as much as in the international context, of a ‘Latin’, natalist and populationist eugenics” (Cassata, 2011, 137).

9However, the decline in the birth-rate was not halted. [5] While in the conviction that the approach adopted was the correct one (Somogyi, 1934), the decline was explained by the action of diverse factors. A number of studies identified a strong correlation between the monthly trend of conceptions (and marriages) and that of the economic-sanitary cycle (Livi, 1931 ; Battara 1931 ; 1935) ; therefore, the failed recovery of the birth-rate was initially attributed to the economic crisis (Livi, 1932). The link between the crisis and fertility decline was not only attributed to the deteriorated economic situations of the families, and therefore the physical conditions of the “procreators”, but was also to the psychological effects that it generated (Livi, 1933). For others, including Benini, the pervasive and increasing reduction in demand for manpower, due to the international economic crisis and the spread of factory work, had made “the population more sensitive to neo-Malthusian propaganda, and spread a real fear of “number” in families. The anti-natalist practices dramatically reduced the birth-rate, which was already in decline in many countries” (Benini, 1936, 85). Lastly, the situation was further complicated by the start of the Italo-Ethiopian War (October 1935) that interfered, at least in the short term, with marriage dynamics and, to a lesser extent, births (ISTAT, 1938).

10In short, however, despite all the arguments presented, the measures of 1927-1928 failed to have their desired effect. [6] The annual trend of the statistics on the birth-rate produced by ISTAT certified the sizeable ineffectiveness of the adopted measures, confirmed by Mussolini himself, who repeatedly acknowledged the failure of the pro-natalist policies adopted by the regime to date in articles and public speeches. In particular, comparing the results obtained in Italy with those in Germany (Treves, 2001, 246 ; Ipsen, 1997, 240, 244), where a series of state interventions in favour of marriage were bringing about a rise in the birth-rate that had reached particularly low levels after the First World War (Livi, 1935a ; 1935b ; 1937).

11Nevertheless, the issue remained central for Mussolini, and so the Gran Consiglio del Fascismo approved an agenda on 3 March, 1937, that was intended to give new impetus to the propulsive population policy. Mussolini’s stance, and the successive measures, were taken on the basis of the consideration that demographers, not only those Italian, had in the past attributed the decline in the birth-rate to psychological factors determined by the crisis, they now largely put it down to “attitudes of free will, determined by a new hedonistic spirit, characteristic of more advanced populations” (Livi, 1938).

12It was therefore decided to intervene on two levels : favouring the marriage rate with the institution of marital loans for couples who married under the age of 26 ; and supporting large families, giving priority to fathers with many working children, while reinforcing pre-existing policies (Ipsen, 1997, 245 ; Provvedimenti demografici, 1937a ; 1937b).

13In this way, the regime intervened in a decisively more substantial manner than previously. Nevertheless, the measures adopted were still far from assuming the dimensions achieved not only in Germany, but also in other European countries. The following year, in Bologna, during a meeting of the CCSP (Comitato di Consulenza per gli Studi sulla Popolazione) attention was drawn to the fact that “the maximum payment made to a family head with six children did not on average exceed 26 percent of his salary, while the corresponding figure was 69.4 and 100 per cent for Belgium and France, respectively” (Parenti, 1938, 416).

14The expenditure of financial and propagandistic energies appeared, at least initially, to have some degree of success, in particular in terms of marriage rates. Even the generic birth-rate showed a slight increase, halting the downward trend. The importance of this change did not escape the demographers of the time (Boldrini, 1940 ; De Vergottini, 1939 ; 1940 ; 1942 ; Giusti, 1941 ; Livi, 1939) who continued to attribute it to the regime’s new policy. [7] However, the modest recovery observed in Italy formed part of a more general trend that affected the old continent and countries overseas – most of all the United States and Australia. At the close of the decade, however, other structural problems presented themselves, not least the arrival at childbearing ages for women from the generations reduced in number in correspondence with World War I birth losses. The feared negative effect on births due to these numerically scarce generations was also aggravated by the depressive effect of World War II that would anyhow have undermined the effects of any policy aimed at incentivizing the number of births. To sum up, therefore, the downward trend in the birth-rate was not halted, even if this decline is to be considered in a political-cultural context that through pushing for its increase could have nevertheless modified its pattern to some degree.

Studies on the decline in the birth-rate in Italy and the 1961 Fertility Survey

15The overall chronology and numbers in the decline in total fertility rates in Italy are well documented on a national and macro-regional level for the period following national unification (1861). A key text of reference on the subject continues to be the work by Massimo Livi Bacci published in 1977, which forms part of a series of national monographs for the Princeton project on fertility transition in Europe. Equally indispensable, for retracing the national evolution of reproductive behaviour, are the period and cohort tables on fertility and marriage rates put forward by various authors between 1968 and 1990 (Livi Bacci et al., 1968 ; Livi Bacci, Santini, 1969 ; Santini, 1974 ; Breschi, 1984 ; Ventisette, 1985, 1986 ; De Simoni, 1989, 1990). This picture is further enhanced with the first regional fertility tables in 1982 (ISTAT, 1982) and those in later years (ISTAT, 1997 and 2000). In short, a rich and varied arsenal of statistical sources is available on an aggregate level to trace the fertility trend of the Italian population at the height of the transitional period.

16If this period is taken as being approximately between the end of the 19^th century and the early 1960’s (Livi Bacci, 1977), the true protagonists of this process, from a longitudinal viewpoint, are these generations born between 1880 and 1925. However, it proves difficult to trace the initial phase of this process with accuracy given that the earliest fertility table dates to 1930, and the first birth cohort with completed fertility was that born in 1915, meaning that a large number of the cohorts that contributed to the births during the Fascist era remain excluded from the analysis. Some light can be shed by data gathered on the reproductive histories of women by the census of 1931, which constituted the first complete survey of female fertility in Italy (ISTAT, 1936). This provided the first opportunity to examine differentials in fertility (Galvani, 1935a, 1935b and 1935c ; Lasorsa, 1936 ; ISTAT, 1936 ; De Vergottini, 1937 ; Battara, 1940) and thus interpret changes in reproductive behaviour underway throughout Italy. However, despite the apparent and propagandistic affirmations on the importance of this innovative survey (ISTAT, 1936), its results were exploited only in part, mainly due to the aversion of the fascist regime to the diffusion of information that exposed clear signs of growing levels of birth control (Corsini, 1967).

17A second, complete fertility survey was re-proposed some thirty years later with the census of 1961. [8] As in 1931, this census gathered information on all ever-married women, but in more detail. However, the Central Institute of Statistics processed this mass of data only in part and with considerable delay, such that the results were not published until 1974. [9] Given that this date fell shortly before the national fertility survey (INF/1) conducted within the wider context of the World Fertility Survey (De Sandre, 1985), which affirmed the superiority of research methods in the field of demography using individual-level data, gathered using sampling procedures as opposed to tabulated data from macro-level surveying, it is hardly surprising that results from the 1961 census received little attention despite its innovative characteristics. [10] While Livi Bacci did make use of this valuable data as part of the Princeton project to highlight the increase in differentials in fertility in women born in the five-year period 1912-16 (Livi Bacci, 1977), the delayed release of these findings also dampened reflections on such ground-breaking studies taken directly from the census returns.

18Our starting point is the “Family sheets of the 10^th General Census of the Population – 15 October 1961” regarding four populations (Turriaco, Novellara, Casalguidi and Alghero) located in four different regions (respectively, Friuli Venezia Giulia, Emilia Romagna, Tuscany and Sardinia). This research was largely conditioned by the availability of, and even more so by accessibility to the original documentation from the census of 1961.

19For the purposes of this study, the relevant information on marriage about each ever-married woman living in the family include the date (month and year) of her most recent/current marriage, the possible date (month and year) of widowhood (spouse’s date of death) or divorce/separation, the dates of birth of parents, and the (start and end) dates of possible previous marriages ; and information (year of birth) on children from current and possible previous marriages.

20These data can then be combined with information regarding the individual woman taken from other sections of the family sheet (such as date and place of birth, education, professional status and field etc.), or put in relation to characteristics of other members of the family group and/or the household’s living conditions (possessory title, number of rooms, presence of bathroom, availability of drinking water, etc.).

21Albeit with a certain degree of approximation, it is possible to reconstruct the complete reproductive history (in terms of intensity and frequency) of the women (married at least once and assessed in the census of 1961) born before 1911, and the more or less complete history of those from the subsequent generations. Given the aims of this study, we therefore have the possibility of carrying out an individual-level analysis on the reproductive history of the women who passed most of their child-bearing age during the fascist era.

22As is typical of retrospective analysis, the Fertility Survey of 1961 also poses problems of selection bias. The principal element of selection is mortality, with high-fertility women being more at risk of dying from complications during pregnancy or childbirth than low-fertility women. This mortality differential would have caused a selection against high-fertility women in the census of 1961, which would also have resulted in an artificially low completed family size for these groups. If we assume that non-educated and low-SES women were those with the highest fertility, any positive evidence of this differential between educational and/or SES groups can be considered a fortiori robust.

23Although processed with notable delay (ISTAT, 1974), the data from the Fertility Survey of 1961 make it possible to reconstruct the reproductive behaviour of the women who reached childbearing age before, during and after the 1930s. Considerable differences in the (marital) reproductive behaviours emerged in relation to the geographical area, socio-economic status and level of education. Again following a cohort-based approach, fertility levels demonstrate a clear downward tendency passing from the women born in the 19^th century to those born in the early 20^th century, even if accompanied by important differences both on a territorial and socio-cultural level (Fig. 1). In Italy, the average number of children per woman fell from 4.4 to 3.3, marking a reduction of more than one child among the women with completed family size.

Fig. 1

Completed (born before 1912) and incomplete (born after 1911) family size of ever-married women (married once and before 45 years of age) recorded at the census of 1961. National, minimum and maximum regional values

Completed (born before 1912) and incomplete (born after 1911) family size of ever-married women (married once and before 45 years of age) recorded at the census of 1961. National, minimum and maximum regional values

24The picture that emerges is less clear when following an approach based on the marital fertility for contemporaries (Fig. 2). The data reported, partially due to estimates for the period 1903-1929 (Livi Bacci et al., 1968), evidently bear the mark of the historical vicissitudes of the first half of the 20^th century. At the start of the century, the levels of marital fertility were relatively high (over 4 children) and in line with those obtained from the Fertility Survey of 1961 (largely corresponding with women born before 1886). The trend remains more or less constant – with a slight downward incline – until the outbreak of the World War I. After the dramatic downturn brought about by the war and flu epidemic of 1918 and the subsequent recovery culminated in 1920, the pre-war levels were not regained and there was a progressive decline “that appears to level out between 1931 and 1940, probably also as a consequence of the propulsive population policy followed by the fascist government” (Livi Bacci et al., 1968, 27). This slight slowing down turns into a momentary recovery in the late 1930s. World War II had a more limited influence on the reproductive behaviour of Italian couples and, also in this case, beyond the brief and rapid recovery, we can observe a substantial stability in the levels at least until 1961.

Fig. 2

Mean number of births per marriage and crude birth-rate, Italy, years 1903-1961

Mean number of births per marriage and crude birth-rate, Italy, years 1903-1961

25The reconstruction of the marital fertility pattern for contemporaries would therefore appear to reveal some effect in relation to the propulsive population policies of the regime. As we have seen, this was also signalled by demographers of the time analyzing the trend of the crude birth-rate that, as we can see in figure 2, closely follows that of marital fertility.

26Given these findings, we propose using the original data of the Fertility Survey of 1961. Using the individual histories of the women from the four communities studied, we are able, with a series of elaborations, to propose an interpretation for contemporaries, which is not possible with the aggregate data processed by ISTAT. This should allow us to measure the possible differential influence on the individual level induced by the fascist policies that, as we have seen, may not have been able to invert the downward trend but could have slowed down the speed of descent.

The four populations studied

27The four populations under study here represent contexts which differ in many respects. One such difference regards population size and population increase (Tab. 1).

Tab. 1

Evolution of the population by the census from 1901 to 1951

Year Turriaco Novellara Casalguidi Alghero N r N r N r N r 1901 1182 7886 4081 10741 1911 1491 0.023 9000 0.013 4165 0.002 11860 0.010 1921 1618 0.007 10156 0.012 4475 0.007 12555 0.006 1931 1778 0.009 10374 0.002 4419 -0.001 13737 0.009 1936 1936 0.017 10637 0.005 4543 0.006 15998 0.030 1951 2378 0.014 11152 0.003 4639 0.001 21374 0.019

Evolution of the population by the census from 1901 to 1951

28The differences are evident and allow us to observe some of their demographic peculiarities at a glance. Turriaco is a community near Trieste that became part of the Italian territory after the First World War. Its population grew notably at the start of the 20^th century, maintained a sustained rate of growth throughout the period and increased even more between 1936 and 1951. In Novellara, in the centre of the Po plain, less than 20 km North of Reggio Emilia, and Casalguidi, an important hamlet of the municipality of Serravalle Pistoiese, the increase slowed down greatly after 1921. Alghero, however, on the north-western coast of Sardinia, after a somewhat contained growth started to rise decisively from 1931 onwards. [11]

29A second major difference concerns the localization and distribution of the population. Whereas in Alghero and Turriaco the overwhelming majority of the population lived in the town centre and its immediate surroundings, in Novellara and Casalguidi a large proportion of people lived in the countryside. At the census of 1951, just after the end of the Second World War, the rural degree [12] was still high not only in Novellara (70.3%) and Casalguidi (53.7%), but also in Alghero (48.8%), despite its urban nature. In Turriaco alone was the majority of the active population involved in labor other than agriculture, in this case shipbuilding.

30Illiteracy is another indicator capable of highlighting the differences between the communities and the profound changes that occurred during the latter part of this time period. In 1951, the illiterate population of 6+ years amounts to 19% in Alghero and tends to decrease with latitude, 15% in Casalguidi, 7% in Novellara and only 2% in Turriaco. Thirty years pre-viously (at the census of 1921) the situation was even more pronounced ; the illiterate population of 6+ was 7% in Turriaco, 21% in Novellara, 14% in Casalguidi and over 35% in Alghero, which provides evidence of the significant educational lag between Sardinia and many areas of Italy at that time.

31Another characteristic which differentiates the four communities, particularly in this historical context, is the political orientations of the majority of the population. [13] While this information is not directly pertinent to this study, it is perhaps worth noting that data from the political elections of 1948 reveal a notable difference between the community of Alghero and the other three. In the mainland communities there was a clear predominance of left-wing parties (over 50%), whereas in Alghero only 14% of votes went to the two main left-wing parties put together and over half to the Catholic party (Democrazia Cristiana).

The analyses of the birth-rate in the four communities

32During the 1930s the birth-rate in Italy reached its maximum level of variability ; geographical areas and social categories that had already undergone the process of demographic transition and reached low levels of fertility coexisted alongside those where natural fertility was still prevalent (Livi Bacci, Breschi, 1990). The situation was also well noted by observers of the time who constantly monitored the situation. In a publication from 1941, Ugo Giusti emphasised this great diversity that is expressed in levels that ranged from 15 to over 35 per thousand from a study of 73 different territorial areas chosen on the basis of certain geographical and socio-economic characteristics. Giusti generalised the different behaviours in two categories : “One that emerges principally in the industrial and urban territories of the North and Centre and that reveals a fairly notable reaction probably to the moral and economic action implemented over recent years to counteract the fall in the birth-rate ; and another, more frequent in the rural areas throughout Italy, that we cannot but consider a sign of insensibility to the above mentioned policy, but that represents the until now negative result between this action and the opposite tendencies connected with the wider socio-economical circumstances in the process of transformation” (Giusti 1941, 117). This variability is also partly mirrored in the four communities under study.

33Figure 3 plots the comparison between the national birth-rate with that of the four communities studied.

Fig. 3

Trend of the birth-rate in Italy and the four communities (1922-1941)

Trend of the birth-rate in Italy and the four communities (1922-1941)

34As we can see, the evolution of the birth-rate during the twenty-year period continued to follow the downward trend already underway prior to World War I. This trend underwent different trajectories in the variegated Italian context. In 1922 the rate was higher than 30 per thousand ; after twenty years, after the start of the war, it was just over 20 per thousand. The exceptions include a small recovery in 1930 and another that appears to be most consistent, at least in terms of duration, from 1938 to 1940. The four communities studied give us a picture that at least partly reflects these characteristics.

35The birth-rate patterns in the four communities, while not in contrast with the national trend, are, as can be seen, extremely diversified. The highest values are those of Alghero, that, even in the inevitable shifting from positive to negative rates, shows a steady trend which even rises at the end of the period. Although with lower levels, akin to the national trend, the birth-rate pattern in Turriaco also manifests a tendency towards growth. Those in Emilia and particularly in Tuscany, on the other hand, show a decisively downward trend.

36The data on these communities refer to populations with diverse structures and behaviour patterns. Only a micro-level analysis can effectively evaluate if there was a change in reproductive behaviour that can be attributed to a new climate that favoured fertility levels. For this reason we devised a logistical model that allows for a more detailed investigation of the relationship between governmental policies and fertility. The Event History Analysis technique was adopted ; a statistical procedure based on longitudinal data that allows for measuring the probability of an event in a certain unit of time – or rather its risk – in relation to a series of both fixed and time-dependent covariates (Allison, 1982). In this case the dependent variable is having a child or not in a given year (Tab. 2).

Tab. 2

Risk of having a child in the year. Completed family size of ever-married women in the four populations

Variables Turriaco Novellara Casalguidi Alghero Freq. OR Freq. OR Freq. OR Freq. OR Age 33.5 0.889 33.4 0.901 33.3 0.903 33.4 0.931 Period (ref. 1930-37) 40.7 1.000 41.9 1.000 41.2 1.000 39.6 1.000 1920-24 14.0 1.810 13.9 1.897 14.6 1.463 16.4 1.302 1925-29 18.4 1.192 19.4 1.312 20.5 1.171 20.8 1.070 1938-41 26.8 0.975 24.9 0.871 23.7 0.939 23.1 0.962 Age at first marriage (ref. <25) 67.6 1.000 71.7 1.000 75.4 1.000 70.5 1.000 25-29 24.6 1.487 21.0 1.467 19.4 1.687 22.4 1.133 30+ 7.8 2.050 7.3 1.689 5.2 2.015 7.2 1.242 Birth place (ref. Province district) 56.4 1.000 87.4 1.000 90.4 1.000 95.8 1.000 Elsewhere 43.6 1.181 12.6 1.209 9.6 1.544 4.2 0.802 Educational qualific. (ref. Primary) 56.3 1.000 70.8 1.000 39.9 1.000 52.1 1.000 None 42.0 1.092 27.9 1.198 59.0 1.270 45.3 1.217 Lower super., Super., University 1.7 0.740 1.4 0.759 1.2 0.522 2.6 0.714 House possession (ref. Own) 75.6 1.000 41.8 1.000 39.7 1.000 39.2 1.000 Rent 20.9 1.415 58.2 1.237 51.6 1.033 60.8 1.045 Unknown 3.5 0.598 - - 8.8 0.985 - - House quality (ref. Low) 8.1 1.000 17.7 1.000 14.0 1.000 6.3 1.000 Medium 76.9 0.901 65.5 0.966 73.5 1.022 89.5 1.324 High 15.1 0.829 16.8 1.006 12.4 0.804 4.3 1.052 Woman-years 4605 19666 9932 6105 Births 770 3435 1547 1656

Risk of having a child in the year. Completed family size of ever-married women in the four populations

Note : in bold parameters significant at p<0.05

37The independent variables of the model take into account certain characteristics of the women and the period under consideration. The time variable has been modulated in a way as to consider the period after 1938 separately, to verify the trend of the fertility rate after the measures introduced by the Gran Consiglio del Fascismo. We therefore adopted the preceding period (1930-37) as the reference category, to form a direct comparison between the years pre and post reform.

38The woman’s level of education is used as a proxy for socio-cultural status. We identified three categories : women with no educational qualifications ; those with a primary school leavers certificate (the reference category) ; those with any higher level of education. Looking more closely at the single socio-professional categories that made up the population, the levels of fertility did not diminish in the same way in all the social classes. The most rapid decline emerged in the classes with the highest level of education. This is a limited number of individuals, never exceeding 3% of the total population. The significance of the variable regarding the level of education is however ambiguous, in that being strongly and positively associated to the economic status of the family, the weight of the cultural component and the economic component are difficult to calculate separately. We therefore decided to introduce certain variables to the model that more specifically control the family’s economic situation, such as ownership and “quality” of the house of residence, in the aim of rendering the interpretation of the education-related variable more markedly cultural and/or behavioural.

39Lastly, there are variables that allow us to scrutinize the demographic behaviours relating to age at marriage and marriage duration and, therefore, their effects on fertility levels. Marital fertility is in fact at its maximum in the first years of marriage and decreases with its duration, independently, within certain limits, from the age of the spouses.

40The results from the model highlight once again the well-known decline of fertility levels. However, this downward trend did ease up after 1937. If pro-natalist policies were not able to invert the fall in birth-rate, they were perhaps able to slow it down.

41This result is by no means surprising, and it was obtained by controlling for many other variables, starting with purely demographic indicators, such as marriage duration and age at marriage. The former clearly demonstrates a progressive fall in the birth-rate with increase in marriage duration, which combines with a rise in the number of births in relation to age at marriage.

42Regarding place of birth, we can see that immigrant women have a tendency to bear more children compared to stable inhabitants, although it is difficult to explain this phenomenon without a more detailed investigation of the places of origin of these women. It should be noted however that at Alghero, where the birth-rate is particularly high, this characteristic is absent.

43When looking at the socio-economic characteristics of the mothers, the strongest sign of birth-rate decline are associated with the categories with higher levels of education. Many of these were white-collar workers, a category that manifested a notably lower average number of children per marriage than the rest of the population (Savorgnan, 1933). On the contrary, the highest levels of fertility (around 20% higher than the reference category) are associated with the categories with no educational qualifications. The only exception in relation to this variable, at least in terms of statistical significance, is given by the community of Turriaco. It is possible that in this case a role is played by the local economy, which unlike for the other communities was predominantly oriented towards the secondary sector.

44The census data allow us to observe particular aspects regarding the connection between type of house ownership and house quality in relation to birth-rate. It emerges that women living in rented accommodation have more children than those living in a self-owned house. If living in 1961 in rented accommodation is taken as an economic marker, and women belonging to this section of the population can be considered as economically disadvantaged compared to the others, we can therefore assert that the poorer socio-economic categories tend to have more children. However, statistical significance is only found in two of the four communities. The data regarding house quality do not give rise to any particular birth-rate differentials except in the case of Alghero, where inhabitants of “average” quality houses have higher birth-rates. Again, an interpretation of these findings would only be possible by further investigation of the particular characteristics of the local environment and residential area.

45Lastly, we have estimated a further model introducing an interaction between woman’s education attainment and period with the aim to check for possible changes across time of fertility differentials by education. However, the increase in model fit (measured through Likelihood-ratio test) was not statistically significant compared to the model without interaction. This means that differentials in fertility by woman’s education level did not differ across birth cohorts.

Conclusion

46Between the late 1920s and the early 1930s, in line with what was happening in numerous European countries, the fascist regime in Italy instigated a series of measures aimed at combating the decline in fertility. Neither the first attempts, that began in 1927 with the Ascension Day Speech, nor the latter measures, adopted in 1937 by the Gran Consiglio’s initiative, obtained their desired effects.

47Despite initial confidence from academic and official sources in the first wave of policies, the subsequent irrefutable results led first scholars and then Mussolini himself to admit that these attempts to revive the birth-rate had failed. Nonetheless, there was a subsequent modest recovery of the birth-rate, at least for a short period which can possibly be partly attributable to the second round of policies activated from 1937 that prompted a modest but significant increase in the marriage rate. However, individual-level analysis of data from 1961 census regarding four communities with different socio-economic characteristics has revealed no increase in the birth-rate within well defined categories of the population. We cannot exclude that these policies had some effect on the choices of single individuals, for example on the basis of the personal compliance to the principles of the regime propaganda. However, even if these behaviours did occur, which cannot be excluded, they do not result as having statistical relevance, at least within the communities examined here, due to the obvious limits in size of our databases. Nevertheless, we should emphasise that while the four communities considered here are not representative of the multiform demographic, social and economic characteristics of the entire country, they are sufficiently differentiated to allow us to affirm that the fascist policies did not result as being more successful from a micro-analytical viewpoint in relation to studies conducted on the aggregate level. This would appear to be in line with a debate that, towards the end of the 1930s, “started circumspectly spreading among certain demographers regarding the effectiveness of the marital loans adopted in Nazi Germany. They were of the opinion that the upturn of the birth-rate in that country was more attributable to recovery from the severe economic crisis, than to the effectiveness of the measures of support for young couples” (Sori, 2002, 111).

48The novelty and size of the rise in births in the four-year period 1937-1940 were well-noted by contemporaries, [14] who had a heightened awareness at the time of anything concerning the birth-rate, and that which was happening elsewhere, such as the German “miracle”, the recovery in Sweden and other European countries and the substantial increase in the birth-rate in the USA and Australia. Upon careful consideration, this increase can be considered – as Anna Treves (2003) acutely observed – an initial episode of the great wave of the “baby boom” in the more than century-long trend of decline in the birth-rate, which was part of an international phenomenon that, in the post war period, had drawn the attention of the demographer Bernardo Colombo (1951). Only recently has a new analysis sprung up around this phenomenon regarding developed countries, prompted by “The fact that the recovery of the birth-rate, and undoubtedly of fertility levels, started prior to or during the war in many countries has been largely neglected” (van Bavel, Reher, 2013). As van Bavel and Reher observe in the conclusions of their study, the extended baby boom (from pre to post WWII) warrants re-examination and re-interpretation. This observation is wholly applicable to Italy, where the question clearly relates to the complex history of the Italian demographic culture during and after the war, entangled in arguments between continuity and discontinuity, interpretative errors and the suppression of past truths held to be inconvenient. This has favoured the creation of a kind of artificial censorship which has, to say the least, misinformed the interpretation of demographic events in their natural continuum. This is a limit to be surpassed with the use of individual-level data, albeit limited, as in this contribution, especially now that the question of policy in the field of population and birth-rate has regained much importance in academic, public and political debate.