METRON - International Journal of Statistics 2011, vol. LXIX, n. 1, pp. 1-28

GIOVANNI MARIA GIORGI

Corrado Gini: the man and the scientist Summary - The paper discusses the times in which Corrado Gini lived in an attempt to enrich our picture of him by adding some further information to his complex personality. Key Words - Scientific research at time of Gini; his presidency of Istat; Gini and Fascism; Gini’s “scientific” ethics.

1. Introduction Even today, 90 years after the first printing of Metron, its name is still closely linked to that of its founder, owner and editor for 45 years: Corrado Gini. When we discussed the possibility of celebrating Metron’s 90th birthday with a special issue, the current editor and I thought that it might be better to minimize the links with its founder. But after looking the question from different points of view we concluded that it is virtually impossible to separate Gini from Metron, the man from the scientist and from the time in which he lived and worked. I have written extensively on Corrado Gini and his scientific work in several articles(1 ). In writing this paper I have tried to take into account some of Samuel Kotz’s suggestions when I wrote some biographies (i.e. Rodolfo Benini, Corrado Gini, Luigi Amoroso and Raffaele D’Addario) for his book with Kleiber(2 ). Kotz suggested that I should also consider the “historical” (1 ) In particular: a Gini’s biography (Giorgi, 2001), a bibliographical portrait of his famous and widely used index of inequality R (Giorgi, 1990, 1993), the relevance and the rediscovery of some of his contributions (Giorgi, 1984, 1999, 2005). I have also investigated methodological issues related to the index R with particular attention to: i) the bounds (Giorgi and Pallini, 1986, 1987a, 1989); ii) the rate of convergence to the limit distribution (Giorgi and Pallini, 1987b, 1990); iii) estimation (Conti and Giorgi, 2001); iv) explicit forms for particular parametric distributions (Giorgi and Nadarajah, 2010); v) the decomposition by sources and subpopulations (Giorgi, 2011). Finally, in a monograph (Giorgi, 1992) I have thoroughly analyzed the huge amount of results reached by numerous scholars on the Gini inequality index. (2 ) Kleiber and Kotz (2003), 261-263 and 265-271.

2

GIOVANNI MARIA GIORGI

times in which leading scholars lived. So this is what I will do in this article on Gini’s life in order to get a more complete representation of the man even if this picture cannot be complete. The relatively recent opening of the private archives and library of Corrado Gini to scholars (see Montevecchi, 2002) and the subsequent publication of a number of historical writings on this important Italian scholar has led me to consider some of his “human qualities” as well as the times in which he lived, especially the Fascist period. In doing so I will also make use of some interviews(3 ) I did a few years ago with two colleagues who knew him and had worked with him, i.e. Carlo Benedetti and Italo Scardovi, along with some conversations I had with Camilo Dagum when I was at the University of Siena. 4)

2. Italian scientific research and statistics at the time of Gini(

Before 1861 Italian scientific research was carried out in many small isolated centers with virtually no interactions between them. As Maiocchi (2003, 14) pointed out, Italy was “largely agricultural, with areas of feudal backwardness” but “with a large number of cultural institutions of higher education and a ratio, between academics and the overall population, among the highest in the world” even though “the level of these institutions was quite different and variable”. A common scientific and cultural basis emerged in the period from the Unification of Italy to World War I. In fact, political and territorial unification facilitated, albeit slowly, the dissemination and merger of ideas among scholars and the improvement of the level of research. In this context Italian scholars “showed distrust for theoretical research, for abstract hypotheses and for excessive mathematization”(5 ). World War I highlighted the need to develop applied research to enhance national capacities and to reduce Italy’s dependence on imported raw materials and products. Thus a so-called scientific-technical nationalism began its development and this continued in the years immediately following the end of the war even if there was a significant slowdown in the early 1920s. After the immediate post-war economic crisis was over, a period of economic liberal-type development began and it lasted until 1929. Scientific research in this period was helped by the founding of the Consiglio Nazionale delle Ricerche (CNR) in 1923 to direct and coordinate Italian (3 ) Giorgi (1996, 2002). (4 ) For a careful and thorough analysis of the subject see Maiocchi (2003) and Pr´evost (2009). In writing this article we made constant reference to some aspects sketched in a recent paper (Giorgi, 2010) and in writings already mentioned. (5 ) Maiocchi (2003, 17).

Corrado Gini: the man and the scientist

3

research. And similarly in 1926 the Istituto Centrale di Statistica (Istat) was established in order to centralize statistical information. But the crisis of 1929 which dramatically highlighted the difficulties of the global economic system, heralded a resurgence of a “scientific-technical nationalism”. This was manifested in increased funding for the CNR which made that agency a powerful tool for mobilizing scientists. But at the same time it was an important component of State organization. In other words, the CNR was used by the Fascist regime to build consensus among scholars and encourage the formation of “militant intellectuals”. So the economic crisis of 1929 was used by the Fascist government to increase its interference in and control over the Italian economy in order to accelerate its project of building a “corporative” society that was considered to be a way out from capitalism’s crisis(6 ). In the years in which the Italian scientific renewal was fervent and “modern” statistics was emerging (i.e. between the early twentieth century and the beginning of the World War I), Gini appeared on the academic scene. From his first appearances at scientific meetings and conferences, his qualities were evident. In a letter dated 29th September 1907, the economist Pasquale Iannaccone wrote to Luigi Einaudi saying that he had noticed “a young man Gini - who is now beginning to work in and with Statistics, but for his culture and seriousness he seems destined to excel soon”(7 ). An excellent prediction, to tell the truth. Gini’s academic career was very rapid. In 1908 he was granted “libera docenza” which gave him access to university teaching. In 1909 he became a temporary professor of Statistics at the University of Cagliari and in 1910, when he was only 26 years old, he was appointed to the Chair of Statistics at the same University. There he also taught Political Economy and founded the first laboratory of Statistics in Italy. In 1913 he took the Chair of Statistics at the University of Padua. But as usual he also taught other subjects like Political Economy, Constitutional Law, Demography and Economic Statistics. At Padua he also founded and directed the Institute of Statistics. In 1925 he moved to the University of Rome (today “Sapienza” University) as professor of Economic Policy and Statistics and from 1927 he was professor of Statistics at the same University also teaching Sociology and Biometrics. In 1936 he founded the Faculty of Statistical, Demographic and Actuarial Sciences. Statistics was defined as “modern” because it was different in form and content from earlier Statistics because it began to use mathematical tools and (6 ) In Fascist aspirations, the purpose of corporatism should have been the establishment a new balance, different from the existing one between the State, society and market in order to avoid the “socialist/communist class struggle” that liberalism had not been able to prevent. For a detailed analysis of Fascist corporatism see Gagliardi (2010). (7 ) See Cassata (2006, 88).

4

GIOVANNI MARIA GIORGI

probability. However, in spite of such developments, researchers continued to distrust its excessive mathematization. To avoid confusion, a line was drawn between the two disciplines. In particular, statisticians discriminated between their Statistics and the Statistics of mathematicians. The first was more practical and the second more abstract. This gave rise to ostracism of Italian statisticians towards mathematicians. In this regard, Pr´evost (2009, 51-52) recalls, inter alia, the criticisms made by the “statistician” Marcello Boldrini (1942) about the book Elementi di Statistica Generale by the “mathematician” Carlo Emilio Bonferroni (1940). It is in this context that we can understand Gini and his followers objections to the Bonferroni concentration index (see, e.g., Giorgi, 1998). Diatribes between statisticians and mathematicians increased the differences and enhanced the separation between the two disciplines. This is confirmed by the statistician Boldrini (1942, 337-339) who noted that the above mentioned book “does not consider Statistics as we understand it, that is as a system of direct methods to solve practical problems”. Boldrini concluded by noting how Bonferroni was at odds with his view of Statistics because, “in the opinion of statisticians, relations are relations between things. What we cannot investigate barely touches us and does not affect us like a mathematician who has an entirely different education than ours”. However, despite these different approaches, the so-called “modern” Statistics was considered as a flexible tool which could be used in a variety of situations and in many different fields. It was not limited to just recording data and taking censuses, but it also had a role in industry, insurance and research laboratories. According to Pr´evost (2009, 64-65), the main contribution to this new way of doing Statistics was the creation of laboratories of Statistics, the first of which was established by Gini at the University of Cagliari in 1910. The use of laboratories seemed to be perfectly appropriate for this new way of doing research where experimental aspects have a significant importance in optimizing domestic resources. Specifically logical-experimental induction was always part of Gini’s research in all the fields in which he carried out investigations. As noted by Fortunati (1942, 361), “even when he theorizes, omitting to write digits [and formulas], in reality, they are constantly present”. In other words, “Gini never forgets to be primarily a statistician, that is a researcher who detects and assesses facts”. Laboratories had a great importance in the development of research because they introduced in research itself and in the teaching of Statistics, methods typical of the so-called hard sciences (e.g. physics, biology, etc.) that were mainly aimed at solving problems. Laboratories and later institutes were places where professors, lecturers and students worked together making better use of human resources and developing scientific knowledge. This new way of organizing work characterized all the research structures directed by Gini.

Corrado Gini: the man and the scientist

5

Division of labour in laboratories facilitated, despite differences between Statistics and Mathematics, the collaboration between statisticians and mathematicians. All this also implied a change in the way of teaching Statistics that became more technical and more practical, less verbal and literary than it used to be. Laboratories and institutes attracted an increasing flow of human and economic resources and enabled the development of statistical publications and increased its visibility. This allowed Statistics to grow and to consolidate itself as a discipline in academia, in government agencies and in other public bodies.

3. Gini and Istat The scientific importance of Gini in Statistics and Applied Economy was so widely recognized that he was appointed to the Consiglio Superiore di Statistica (CSS) in 1911, and in 1926 he created the Istituto Centrale di Statistica (Istat) - a centralized structure for the collection of official statistical information. In fact, until then, this activity was fragmented between Ministries and local governments without a real and effective osmosis of information between various agencies. From 1926 to 1932 Gini was President(8 ) of Istat. During that period he organized and coordinated national statistical services and took them to high efficiency levels in spite of the difficulties involved. In fact, the obstacles created by senior bureaucrats in various Ministries and the passive resistance of most public agencies to his work of centralizing statistical information gave rise a lot of dispute that Mussolini was frequently called to settle. This ruined the good relationship that Gini had initially established with Mussolini who was attracted and interested in Statistics and its relevance to State affairs. Mussolini had understood the propaganda power of numbers and therefore of statistics which could be used for exalting the results achieved by his regime and for masking its failures. He considered “statistics as a tool to be used for policy and to forge the idea of nation”(9 ). Because of his great interest in statistical information and particularly in demographic data, the Duce was (almost) in daily contact with the President of Istat in order to be continually updated not only on births, deaths, marriages in different regions and in major Italian cities, but also on economic data related to agricultural and industrial production and to imports and exports. The data were (8 ) In 1926 Gini was appointed President of CSS and not of Istat because the law had not yet provided this office. However, he was named President in the inaugural speech and he signed all correspondence as President from that day on. Only in 1929 a decree-law introduced the position of President of Istat. On this topic see Leti (1996, 108). (9 ) Misiani (2007, 53). For the relationship between official statistics and policy in those years see also Favero (2004, 49-59).

6

GIOVANNI MARIA GIORGI

related to both Italy and other countries like Germany, England and France. However, he did not just ask for information but he criticized, suggested and even intervened in the details of data production. Sometimes Mussolini by-passed Istat by directly urging prefects and mayors to publish and update statistical data and to solicit directors of newspapers to publish data on marriages and births that, if different from the goals and expectations of the Fascist regime, were severely criticised in order to shake-up the “bourgeois” apathy of citizens(10 ). Treves (2001, 215) ironically comments that Mussolini believed in statistical data “with the same intensity and conviction with which many politicians believe in opinion polls today”. Over the years, the interference of the Fascist regime became more and more frequent. It was a clear signal that “the times were changing” and that “Gini’s independent and impatient spirit would brook no interference with his work”(11 ). Gini understood that his special relation with the Duce, without which he would not have had the necessary support for his work, was rapidly disintegrating. He concluded that it was better to leave the presidency of Istat before the controversy about the organization of the national statistical service evolved into power conflicts which would have done serious harm to his work. So he resigned and in his last service order on February 10, 1932 he stated that, after having ended his work at Istat, he would return to his studies. According to Castellano (1996, 12), this decision avoided “the danger of a political succession to the Istat presidency, that the involution of the system could have made a reality”. The successor was professor Franco Savorgnan whose personality was more flexible than Gini’s. Cassata (2006, 13) suggests that among the possible causes of Gini’s resignation, we should also consider the different points of view of Mussolini and Gini on the relationship between science and politics. Mussolini thought that technicians should provide “only contributions in terms of intellectual guarantees or simply of image”. While Gini believed in the more important involvement of technicians in preparing, interpreting and managing statistical data. In other words, Gini was a defender of scientific autonomy and therefore faithful to “the criterion according to which an intellectual would assess events and political decisions on the basis of his own schemes and theories”(12 ). With regard to the period when Gini was President of Istat, although he respected the political power, he never actually submitted to it, almost always stressed independence and freedom of opinion. Indeed, it could be said that he used political power to improve the organization and the reliability of the national statistical service. (10 ) For an interesting analysis of the relationship between Mussolini and the Italian official statistics, see Leti (1996, 475-539). (11 ) Castellano (1965, 7). (12 ) Treves (2001, 227).

Corrado Gini: the man and the scientist

7

In particular, according to Leti (1996, 554), “the impression that one has from the [Gini-Mussolini] correspondence is that of a man who conversed with a powerful dictator for his work but did not submit to Mussolini”. These views are shared, at least in part, by other scholars. Thus according to Trivellato (2004, 78-79), Gini’s participation in Fascist policies and particularly to that inherent the population and official statistics showed “significant independence of judgment” based on his convictions and awareness of his own caliber. In other words, there was “very little servility in Gini; indeed, he often proudly affirmed his ideas”. Nevertheless, we must not forget the “nearness of some of his beliefs to Fascism and to the authoritarian regimes of the time (apart from racial policy)”. According to Castellano (1996, 12), who was the closest collaborator of Gini in those years, the involution of the regime was perceived by Gini “with some delay” (as other intellectuals of the time) and hence “in all its seriousness for him, Istat, the Nation and his dream of a Faculty [of Statistics]”. With resignation, his disengagement began. But he suspected that the political battle against him was only postponed and that there would be many obstacles along the way to the achievement of other important goals. His fear was justified because he well knew what the results of the delation were since there had already been several complaints against him to the political police and he had been the object of a hostile press campaign(13 ). Cassata (2006, p. 93-97) mentions that these complaints were made directly or indirectly between 1927 and 1931 by those who wanted to be employed at Istat, only because they were members of the Fascist party. They tried to discredit(14 ) Gini by accusing him of being overly authoritarian and of acting in defiance of Fascist law because he had put “subversive” elements like the socialists Molinari (as general manager) and Silvestri (the chief of Department III) and non-party members like Arcucci (chief of staff) into key positions at Istat. However, throughout the period in which he was president of Istat, Gini opposed the pressures of the regime with regard to the management of staff by motivating the recruitment of non-Fascist officials and even socialists for their organizational skills and knowledge of modern statistical methods. In this context, particular attention should be paid to Alessandro Molinari, the general manager of Istat and Gini’s main collaborator. Formerly, Molinari was the director of the Statistical Office of the municipality of Milan and was chosen by Gini for his excellent organizational skills. Misiani (2007, 53-59) in his accurate reconstruction of the recruitment of Molinari to Istat, shows that Gini had been in contact with him since 1927, (13 ) The press campaign was orchestrated by the regime following an interview by Gini concerning the sinking of the liner “Principessa Mafalda” on which he had embarked. For more information on the topic see Cassata (2006, 96). (14 ) On the subject see also Leti (1996, 529-535).

8

GIOVANNI MARIA GIORGI

but that Molinari was not very interested in accepting Gini’s offer because he did not want to become involved with Fascism. However, in spite of his initial refusal, talks continued - with Mussolini’s approval - between Gini and Molinari for a couple of years. In the end Molinari surrendered to the insistence and charisma of Gini and he took office at Istat in late 1929. This makes us wonder why the Duce endorsed the recruitment of nonfascists and/or socialist scientists and technicians and even more why Molinari accepted Gini’s offer. Historians seem to agree that Mussolini was in favour of greater cooperation with the non-Fascist technicians in the hope, as Misiani (2007, 60) claims, “to be able to influence their actions and move them towards Fascist conservatism in order to strengthen his totalitarian political project”. All this was part of the general aim of Mussolini’s policies with regard to Italian culture and its “fascistization” which began between 1925 and 1926. In fact Alfredo Rocco’s laws, enacted after the Zaniboni assassination attempt on Mussolini, marked the transition from the “liberal” to the Fascist State(15 ). In those years Fascism began “to invest” in Italian culture. The first step was the “Fascist intellectuals manifesto”, then the constitution of the “Fascist National Institute of Culture” and the foundation of the “Italian Academy” that was joined, among others, by important scientists such as Guglielmo Marconi and Enrico Fermi. Sedita (2010, 31) recounts that members of the Italian Academy were entitled to an annual allowance of 36,000 Italian lire plus other allowances for particular positions established by the Academy itself. However, it was in the period between 1930-1940 that the regime made its greatest effort to tie the Italian culture to Fascism by allocating “more than 600 million Italian lire to fund over 900 intellectuals and 400 newspapers”. To emphasize the importance of the amount of money allocated by the government to this project, Sedita (2010, 17) notes that “in 1929 (before the Great Depression) the per capita income of Italians was 3,079 lire per year”. Let us now consider the reasons that prompted prominent personalities, scientists and technicians to participate in some aspects of the Fascist regime’s policies - among which the autarky was one of the most important. With regard to this Maiocchi (2003, 125) notes that in addition to the usual opportunists, there were “many people who considered the autarky as a real opportunity for the rationalization and progress of Italian manufacturing”. So scientist by giving their contributions to scientific-technical nationalism could receive new opportunities for research and the appropriate monetary compensation. Therefore the participation in autarkic policy also included personalities adverse (15 ) For a careful analysis of the “liberal” phase of fascism see Perfetti (2010, 19-27) and in particular Cassese (2010) who by using a juridical-institutional approach, shows a substantial continuity between Fascism and the previous liberal period, pointing out that Fascism used pre-fascist institutions and their technical-political staff extensively.

Corrado Gini: the man and the scientist

9

to the regime. In doing so they became, albeit indirectly, supporters of the Fascist regime’s policies. In fact, their studies and their research did give “coverage of scientific-technical type to the ideas of the government”(16 ). So with regard to Molinari - never enrolled in the National Fascist party - his cooperation remained purely technical and, as Misiani (2007, 61) claims, “was not a surrender to the flatteries of the political power [...] but should be interpreted as an attempt to influence the regime from within”. Following Misiani’s line of argument (p. 67), one might say that “Molinari was opposed to the idea of Official Statistics in service to the regime, but was favourable to the idea of a pragmatic and empirical understanding of government policy which entrusts the statistician with the role of providing knowledge of the actual situation on which to base public policies”. At Istat Gini introduced an organization based on technical expertise and research actuated by recruitment and career advancement that took account of scientific and technical merits rather than seniority. He also favoured interaction between universities and research centers and he encouraged participation in international meetings in an attempt to make Italian official Statistics less provincial. We can see from the interpretation of the results of the 1921 census, which showed a falling birth rate, that Gini and his general manager Molinari considered statisticians as independent figures who interpreted the results of research and not as executors of regime’s policies. These results did not meet the expectations of the government that, of course implemented a policy to increase birth rates. In other words, census data showed that, in contrast to regime’s goals, there was not a link between the country’s development and population growth and indeed the data showed a clear relationship between production dynamics, urbanization and the reduction of births(17 ). Gini’s scientific and professional ethics led him to verify the accuracy of these results. So in February 1928, this verification was deepened by analyzing a large number of municipalities that had been shown a decrease in the resident populations between the censuses of 1911 and 1921(18 ). According to Molinari (1932), the main reasons which caused this demographic trend were: i) an increase in economic well-being and the desire to maintain and/or increase it; ii) an increase in economic hardship and the desire to reduce it; iii) increased industrialization; iv) the growth in family expectations towards a better life for themselves or their children; v) a physiological decline in fertility. Obviously the Fascist regime hesitated in the publication of these results because they “contradicted the fundamental elements of its ideological design”. (16 ) Maiocchi (2003, 309). (17 ) Misiani (2007, 68). (18 ) Molinari (1932).

10

GIOVANNI MARIA GIORGI

Although a staunch supporter of the regime’s birth policy, Gini remained faithful to the principle of research autonomy from political influences. Thus, in order to break the deadlock he instructed Molinari to present these findings to the International Conference on Population held in Rome in September 1931. It seems that Molinari spoke from a personal point of view(19 ) in order to avoid Istat’s direct involvement. However, Gini’s support for the “true” results of the survey was evident both because, as President of the Italian Committee for the Study of Population Problems (Cisp), he organized the conference and also because he published Molinari’s report under the title of “An investigation into the reasons for the falling birth rate” in Metron (1932) - an international journal edited and owned by Gini himself. As mentioned above, Gini also founded the Faculty of Statistics in 1936. According to Federici (2000, 18) the main goal of the Faculty was, as Gini saw it, the production of university graduates who were able to combine theory with practice “in the belief [...] that all governments should recruit statistical officials for assisting them in their political decisions”. The Faculty was the natural evolution of the School of Statistics founded in 1927 from the merger of the School of Statistics and Actuarial Sciences - established in 1926 at the Faculty of Sciences by the mathematicians Guido Castelnuovo and Francesco Paolo Cantelli - and Gini’s Institute of Statistics and Economic Policy.

4. Gini and Fascism We have seen that it is virtually impossible to deal with Gini and his relationship with the University and Istat without referring in some way to Fascism. On the other hand, like all dictatorial regimes, Italian Fascism was like an octopus with its tentacles everywhere in an attempt to keep everything under its control. Therefore, in order to make the reader aware of the historical period in which the scientist Corrado Gini lived and worked, as well as his relationship with Fascism, it must be remembered, albeit briefly, that in 1925 Gini had signed the “Fascist intellectuals manifesto” and was one of 18 “wise men” that Mussolini commissioned to prepare the Fascist Constitution. Further, it should not be forgotten that he was the author(20 ) of “The scientific basis of Fascism”. In other words, he was not just a “non-opponent” of the regime, like many of his colleagues. In fact, it must be remembered that at that time, self-interest and conformism played important roles in academic life and in most aspects of Italian life. These tendencies can be inferred from the fact that, in 1931, (19 ) Misiani (2007, 68). (20 ) Gini (1927).

Corrado Gini: the man and the scientist

11

only a dozen out of 1200 professors refused to swear allegiance to Fascism and as a result they lost the right to teach at the University. However, as Sergio Romano writes (2006, 39), prominent figures advised the university professors to sign the oath of allegiance for various reasons. Thus, for example, “the Pope [Pius XI] informed Catholic professors that they could sign the oath with so-called ‘mental reservations’. Togliatti [the head of the Italian Communist party] argued that they would have been more useful to the party if they had stayed at the University. Similarly the philosopher Croce urged professors not to interrupt their ‘liberal’ teaching”(21 ). Even Enrico Fermi, Nobel Prize for Physics in 1938, for whom financial support from the Fascist regime was crucial to his research, had been a member of the National Fascist party since 1929. Fermi did not openly oppose to the regime, at least as far as we know, until he was forced to leave Italy because of the 1938 racial laws which affected his wife who was Jewish. As Paolo Mieli (2001, 41) reminds us, “Fermi, as with many other Italian intellectuals [...], discovered the oppressive and totalitarian nature of fascism only when he was personally hit”. It was in this context that it is also worth considering what happened to the statistician/demographer Giorgio Mortara. According to Baffigi and Magnani (2009, 252), Mortara perceived, like many other Italians (Jews, Christians and Atheists included), “the beginnings of the Fascist racial policy as aiming at ‘improving’ the Italian race but not as an incipiently anti-Semitism”. Mortara had joined the Fascist party(22 ) in 1933 and, despite this, he realized, like Fermi, what the real purpose of the racial laws was only when they affected him and his family. However, except for a small minority, Italians adapted themselves to this “racist tendency”. Although many secretly disapproved these laws, most limited their dissent to attitudes of individual piety. Gini, in his own way and with great prudence, considered the racial laws of 1938 “significantly at odds with his vision of society and his willingness to favour interbreeding between populations”(23 ). His ideas, with regard to this, were clear before the advent of Fascism and, given the nature of the man, he certainly did not change them later. For instance, he had been in favour of racial interbreeding since the beginning of his academic career, as he stressed in his book (Gini, 1912) “I fattori demografici dell’evoluzione delle nazioni”. In some cases, as remarked by Pr´evost (2009, 98), he spoke of Fascism’s aberrations particularly in the field of racism and this is why he disbanded the Editorial (21 ) See also De Sandre (2004, 51). (22 ) Mortara’s membership in the Fascist party was not an anomaly because according to some scholars (see Baffigi and Magnani (2009, 244)), “the degree of adherence to fascism of the Jews was similar to that of other Italians, although the proportion of Jewish anti-fascists was higher than the overall national proportion”. (23 ) Trivellato (2004, 78).

12

GIOVANNI MARIA GIORGI

Committee of Metron(24 ) in order to avoid the radiation of the Jewish members imposed by the regime. Nevertheless, Jean-Guy Pr´evost (2009, 250) also notes that Gini delayed his moderate opposition to some choices of the Fascist regime until after the fall of Mussolini. But this did not prevent him from being impeached in the purge processes(25 ) that took place from the summer of 1944 within the Public Administration and the University. Thus, on 6th November 1944, while awaiting the results of this process, he had to leave teaching and also his position as Dean of the Faculty of Statistical Sciences which was temporarily given to Filippo Vassali (Dean of the Faculty of Law) and Roberto Almagi`a(26 ). In that period, he also had to leave the Presidency of the Italian Statistical Society that he had occupied since 1941. Pr´evost also comments that - although acquitted of the most serious charges on 24th January 1945 - Gini was suspended from all academic duties and his salary was not paid for one year. Gini and the High Commissioner in charge of the prosecution appealed, of course for opposite reasons, against this ruling, but the latter submitted his reasons after the deadline and the subsequent act of 17th December 1945 was “not to prosecute”. However, Gini survived virtually unscathed from that period and in 1946 he resumed his duties at the Faculty and in 1949 was again the President of the Italian Statistical Society, a position he held until his death. Certainly, his departure from politics was almost complete and the last twenty years of his life were mainly devoted to his studies. In 1955 he was appointed Professor Emeritus at the University of Rome. In any case it is important to note that there was no uniformity of treatment in the purge processes. For instance, a number of supporters of the racial laws, who benefited from their application, did not pay for their actions and remained in their chairs and positions of power in academia. In particular, Ventura (1996, 179-180) notes that Gaetano Pietra - professor of Statistics and one of Gini’s collaborators and his successor at the University of Padua - was a supporter of regime’s racial policies. In fact not only did he continue to give seminars and conferences on questions of race, but he also assumed the teaching of a course of “General Demography and Comparative Demography of the Races” - a new university course introduced by the Ministry of Education under the racial laws. He was also appointed “Dean of the Faculty of Political Science [...] to replace Donato Donati deposed on grounds of race”. And subsequently he became the Dean of the Faculty of Law (1941-1943). Moreover, Ventura stresses that Pietra “consistent with his anti-Semitic commitment, hoped that, (24 ) The names of the Editorial Committee were removed from the cover of Metron from the issues 1 and 2-3 of volume XIV published in 1940 and 1941. In addition, Gini never allowed the years of the “Fascist era” to be printed on the cover of Metron, unlike other scientific journals. (25 ) For a detailed and thorough analysis of the purge process against Gini, see Cassata (2004). (26 ) See Donati (1996, 9).

Corrado Gini: the man and the scientist

13

because of Fascist racial measures, land ownership could be changed in favor of peasants. This proposal involved the confiscation of Jewish property”. However, despite all this, Gaetano Pietra became, after the liberation from Nazi-Fascism, first a Commissioner, then the President of the Camera di Commercio of Udine and he was also elected as a Senator of the Christian Democratic party in 1948(27 ). More recently, Cocchi and Favero (2009) have highlighted the fact that several of Gini’s contemporaries, who had openly favoured the racial policy, suffered no consequences. Having outlined Gini’s complex relationship with Fascism and how this impacted on him, we would like to refer readers interested in this subject to the works quoted in the footnote(28 ).

5. Some Considerations on Particular Aspects of Gini’s Scientific Work Corrado Gini investigated many topics, driven as he was by a boundless curiosity and his desire to apply Statistics to many practical problems. That Economic Statistics was among his favorite topics can be clearly inferred from the opening address(29 ) to the course of “Economic Policy and Statistics”. In fact, when he was called to the chair of Economic Policy at the University of Rome, he “wanted Economic Statistics to be included” because it provided “the basic information necessary for Economic Policy studies”. Already in 1910, he published a research(30 ), in Giornale degli Economisti, aimed at analyzing the relationship between prices and consumption. It was a pioneering work for that time when Econometrics had not yet become a subject in its own right(31 ). This kind of interest lead him to deal with the theory of (27 ) For further information on Pietra see Favero and Trivellato (2009). (28 ) In particular, see Favero (2004) on Gini and Italian Statistics during the twenty years of Fascism; Pr´evost (2001) on Gini and Liberal Democracy; Bertaux (1999) on Gini and his relationship between science and ideology; the volume edited by Dalla Zuanna (2004) on Statistics and Demography in Italian culture between the two World Wars. We also recommend the book by Giuseppe Leti (1996) on Istat and the Consiglio Superiore di Statistica and Jean-Guy Pr´evost’s (2009) excellent monograph which is a competent and objective study of Italian Statistics during Fascism. Finally, Anna Treves’ (2001) monograph with particular reference to the chapters on the question of births during the Fascist period, and Francesco Cassata’s (2006) volume that investigates, through his biographical sketch of Gini, the links between science and politics in order to assess the level of autonomy that existed between politicians and intellectuals in the Fascist period. (29 ) Gini (1926a, 137). (30 ) Gini (1910). (31 ) The Econometric Society was founded by Irving Fisher and Ragnar Frish and the term “Econometrics” was introduced in the literature by the latter in 1926. However, as argued by Georgescu-Roegen (1966, 265), Gini “did not have to wait for the Econometric Society to be founded. He had been an ‘econometrician’ in the true sense of the word for a long time before this term was coined”. On this subject see also Roy (1966, 283-286).

14

GIOVANNI MARIA GIORGI

index numbers to which he gave significant contributions(32 ). Thus he was able to construct a solid bridge between statistical theory and economic theory through the so-called “Gini identities” as named by Ragnar Frisch (1936). Later, Benedetti (1984) showed that the Kon¨us index was a direct consequence of these identities. Recently, Cuthbert (2003) confirmed that the aggregation method developed in 1964 by Eltet¨o and by K¨oves and Szulc had already been anticipated by Gini forty years before(33 ). However, studies on income and wealth(34 ), especially those related to the measurement of the distributive inequality of these variables, were a success that gave Gini international fame. From the early years of his scientific career, Gini also did research into questions concerning wealth. He concentrated on several of its aspects by conducting a comparative study of the qualitative composition of wealth and by establishing a picture of its dynamics in various countries. He also highlighted the need to include the evaluation of human capital in the computation of wealth, a need reiterated in his research on the causes and consequences of international migration. According to Nora Federici (1960, 15), it was precisely the consideration of human capital that allowed Gini to “interpret in an absolutely new way the causes of American prosperity, by documenting how this prosperity should be considered as mainly determined by the contribution of wealth that Europe had provided to the New World in the form of huge masses of emigrants”. Wealth and income variables were often the subject of many of Gini’s studies and writings that were collected in a volume(35 ) in 1958. In addition, he was responsible for a number of initiatives(36 ) in the national and international field. For example he had such a good reputation that in 1922 the League of Nations commissioned him to assess the income and wealth of its member States. In 1931 the International Statistical Institute (ISI) in its Tokyo session, awarded Gini the presidency of the newly-formed Committee for the study of wealth and income and therefore with the task of preparing a preliminary report for the ISI session of London in 1934. After World War II, in 1947, he presented the National Commission for Income of the National Economic Council, a comprehensive report on “content and use of national income assessments” which was adopted unanimously. In 1948 in Rome, Gini also founded the Italian Committee for the Study of Income and Wealth and (32 ) Gini (1924). (33 ) See also Biggeri et al. (1987, 569). (34 ) According to some scholars (see S. Lanaro quoted in D’Autilia (1992, 52-53 footnote 9) for Gini and his other collaborators, “the calculation of national wealth [...] is never a purpose in itself: it should rather serve to identify the social dividend and growth factors to orient toward objectives that measure the reproduction of capitalism”. (35 ) Gini (1958). (36 ) See Pedroni (1960, 516-517).

Corrado Gini: the man and the scientist

15

in 1949 he participated in the establishment of the International association for Research on Income and Wealth. Studies by Gini on the measurement of inequality (or concentration, as he preferred to say) were hotly debated within the political-economic academic community. This debate - very lively in the late nineteenth century - was concerned with achieving a more equitable distribution of income and wealth. It was in this context that Gini (1909) first proposed the so-called δ index and then in 1914 the concentration ratio R. There has been considerable scientific production on these topics(37 ). Here we limit ourselves to remember just the important contributions to the decomposition of R that continue to be discussed(38 ). New decompositions make R more and more operational and modern, thus providing technicians and policy makers with a tool that takes into account aspects of great importance in today’s societies. Gini’s many theoretical and methodological contributions have successfully passed the test of time by maintaining an enviable and sometimes inexplicable topicality. What is interesting is that he made these contributions by relying as little as possible to sophisticated mathematical tools. He confronted reality by trying to understand the logic behind the events and only after he used those statistical and mathematical tools strictly necessary to make a model or - as he preferred to say(39 )- a scheme. He regarded this latter as “a more appropriate term because it reflects the simplification of reality that the process involves and because it contains no suggestion of the normative character that, at least in some languages, the word ‘model’ implies ”. Moreover, according to Gini “the essential condition for making progress through models is that their results should be verified. Since models (or schemes) present a simplified or often simplistic reality, they must be trusted only if the simplifying assumptions do not imply that they lead to conclusions too far from the facts. As soon as new facts emerge, a model check must be conducted” to avoid “models multiplying, overlapping and often contradicting themselves”. According to Gini a thrifty and careful use of mathematics in Statistics was the modus operandi to prevent the development of complicated structures which only obey their internal grammar and are difficult to use in practice. At least in part this point of view was shared by other scholars outside of his entourage who tried to combat the perverse use of methods that are just complicated mathematical exercises and in practice difficult to use. Along the same line, Georgescu-Roegen wrote(40 ) that “because of the well-known difficulties of getting down to empirical brass tacks in economics, many a student has found it more comfortable to continue the formalism of mathematics (37 ) See references quoted at the beginning of this paper. (38 ) See Giorgi (2011). (39 ) Gini (1956, 175-176). (40 ) Georgescu-Roegen (1966, 266).

16

GIOVANNI MARIA GIORGI

or, as this is often put, to substitute mathematical exercises for economics”. Similarly Amartya Sen (1999, xxv) is on the same wavelength when he assesses various ways of decomposing an inequality index and he insists that it is not sufficient to simply consider the formal aspects. In particular, he is referring to entropy-based measures like the Theil index which merely satisfies its internal consistency - represented by the additive decomposability - without taking into account other important aspects of inequality. Therefore, we can say that Gini’s scientific production was rarely the result of just a theoretical-formal elaboration but that it arose from the need to solve concrete problems. He was able to affront these problems from different angles because of his extensive and profound knowledge of various fields such as Economics, Biology, Sociology and Demography. His methodological contributions relate mainly to the analysis and comparison of statistical indices for the study of the variables from a population rather than a sample. In broad terms, his purpose was to investigate mean values, variability, concentration and the association of statistical variables. We have extensively discussed on other occasions some of these topics. Here, we simply observe that other important aspects of his work concern “averages” (Gini, 1938) which he classified by their analytical aspects while also providing general formulas. He also studied problems relating to the extension of the concept of mean to categorical variables. The results of over twenty years of research on “averages” were collected in a book (Gini, 1957a). Recently, after decades of indifference, mathematicians have again started to study some properties and relations of the so-called “Gini means”(41 ). Nevertheless we must not forget Gini’s interest in statistical inference which highlights some features of his personality. His interest in this topic can be traced back to his graduation(42 ). But it was only in 1911 that he decided to investigate this topic in depth when, in order to make inference on the relationship between human genders at birth, he began to look at the use of the Beta probability distribution as an a priori law(43 ). However, it was during the decade 1939-1949 that he devoted special attention and renewed interest to the clarification and analysis of inferential problems. On the occasion of the first scientific meeting of the Italian Statistical Society held in Pisa on 9th October 1939. In his inaugural address, entitled “The Dangers of Statistics”, Gini criticized Fisher’s fiducial methods and the Neyman and Pearson framework for testing hypotheses because he considered them weak from a logical point of view(44 ). With regard to this topic, some scholars have stressed the contrapo(41 ) See, e.g., Farnsworth and Orr (1986), Neuman and S´andor (2003), Czinder and P´ales (2005), Shi et al. (2009), Baj´ak and P´ales (2009). (42 ) See Gini (1908). (43 ) Gini (1911). (44 ) Gini (1939).

Corrado Gini: the man and the scientist

17

sition between the Italian and the Anglo-Saxon statistical schools. When Gini spoke about the Italian statistical school and Italian statisticians, he was actually referring to himself(45 ) and in this case he was making a comparison of his ideas and his view of the theory of inference with those of Fisher and Neyman and Pearson. His attitude was that of a scholar regardless of his nationality. It is also worth remembering that Gini had a bitter dispute with Bortkiewicz(46 ) in 1931 about the originality of the contributions of Italian statisticians(47 ) and the authorship of some scientific results. It also needs to be remembered that Gini was always convinced of his own abilities and of the logical correctness of his reasoning. This attitude often caused him problems. According to Treves (2001, 227), this was due to “ the overflowing certainty he had of the high value of his thought”. An example of this somewhat arrogant attitude is his criticism of the theory of inference and his absolute conviction, at least in the beginning, that he was not only right but that his ideas were completely original. So that at the 7th Scientific Meeting of the Italian Statistical Society held in Rome on 27-30th June 1943, he stated: “Nor is there any wonder that my criticism has caused an uproar in the world of statisticians, many of whom have the impression that the whole ‘building’ of our discipline has been shaken at its foundations, and others that my criticisms could make it collapse”(48 ). This conviction, as noted by Forcina (1982, 65), was in part due to the fact that “Gini’s knowledge of the work of Fisher and that of Neyman and Pearson, after close examination does not seem to be entirely adequate”. The same opinion is also shared by Herzel and Leti (1977, 26) who, after a thorough and lucid analysis of Gini’s criticism to statistical inference, point out that some of his positions were the result of a superficial interpretation of Fisher and Neyman and Pearson’s theories. Nevertheless, according Herzel and Leti, Gini’s critical position generated a debate that resulted in a benefit, albeit indirectly, for the theories concerned because it contributed to a clarification of their assumptions. In this Metron special issue, Piccinato (2011) - in an intelligent re-reading of some writings on the subject - highlights some other weaknesses of Gini’s (45 ) See also Treves (2001, 227). (46 ) See Bortkiewicz (1931a,b) and Gini (1931a,b). Gini’s positions were also supported by his followers Pietra (1931a,b) and Savorgnan (1931). (47 ) However, it should be noted that this claim was not merely the result of Fascist nationalist rhetoric and of self-promotion of the dominant group. It actually corresponded to the real scientific dynamism of the Gini group (Cassata, 2006, 144). Pr´evost (2009, 163) also believes that these claims had a concrete base and highlights some objective evidence like the inclusion of an appendix on Italian statistical methodology in various editions of Yule and Kendall’s “Introduction to the Theory of Statistics”. Several dozen of terms translated from Italian that did not have equivalent words in Anglo-Saxon Statistics were included in the first four editions of Kendall and Buckland’s “Dictionary of Statistical Terms”. (48 ) Gini (1943, 245).

18

GIOVANNI MARIA GIORGI

arguments and also the attempt by some scholars (e.g. Pompilj) to reconcile Gini’s position with modern statistical inference. From the above it is clear that in Gini’s opinion he was the Italian Statistics!(49 ) On the other hand, at that time many of the most important results in Statistics were achieved mainly by him and by researchers who had gathered around him. Also the Italian Statistical Society - founded in 1939 initially chaired by his follower Gaetano Pietra - disseminated, almost exclusively, the results of Gini’s research and those of his collaborators. In other words, at that time in Italy, there was no real competition of ideas in the field of Statistics. For over half a century, Gini was the dominus of Italian Statistics, i.e. the one who imprinted scientific production by suggesting the research topics to be developed. The work of his collaborators, though innovative and original, was never truly independent. They always followed in the path traced by their master, i.e. they developed, investigated and elaborated what Gini had “sown” and suggested. This was the situation in most cases and at least until the late 1950s or more specifically until Gini’s retirement. However, the originality and modernity of some of his contributions and those of his collaborators are undisputable and the validity of some of his less known contributions are being rediscovered even at an international level(50 ). 6. Some Concluding Remarks The literature shows that Gini’s scientific production on statistical methodology is still topical and the recent acquisition of his library and archives by the Archivio Centrale dello Stato has thrown the first rays of light on some aspects of his life. However the results of his studies are still of great interest for international scientific community and many articles, both theoretical ad applied, on the subjects treated by Gini continue to be published in journals of various disciplines from Econometrics, Applied Economics, Economic Statistics, Mathematical Statistics to Informetrics, Bibliometrics, Ecology, Medicine, Biology and so on. These papers consider extensions, interpretations, insights not only (49 ) Even Giorgio Mortara, in exile on racial grounds, considered Gini “a scholar worthy of utmost respect for making the Italian statistical school appreciated abroad, even if he made it appear as if it consisted of himself and his entourage” (Baffigi and Magnani (2009, 250). (50 ) See Giorgi (2005) and also Montanari and Monari (2008), who have highlighted the impact of some results by Gini on multivariate statistics. In particular, on classification and regression trees, multiple linear regression, principal components and linear discriminant analysis. Further, we have to remember Yitzhaki’s efforts (with other authors) to rewrite some statistical topics by using the so-called Gini method which is composed of statistics known as Gini covariance, Gini correlation, Gini regression coefficient etc. On this subject see Yitzhaki (1991), Yitzhaki and Olkin (1991), Olkin and Yitzhaki (1992), Schechtman and Yitzhaki (1987, 1999, 2003), Frick et al. (2006), Schechtman et al. (2011).

Corrado Gini: the man and the scientist

19

on the most famous of Gini’s issues, but also those less known internationally that are being, slowly but steadily, rediscovered. The continuous rediscovery of known results is due to the difficulty that many non-Italian researchers had and have of accessing the original works of Gini and of other important statisticians of the Italian tradition. This difficulty is due to the fact that Gini and his scientific followers wrote almost exclusively in Italian and quite often in journals of limited international distribution. While the occasional summary(51 ) of their findings made by Gini(52 ) - in English or other languages at international meetings and also in prestigious journals - were not enough to allow widespread and detailed knowledge of what Gini and the Italian School of Statistics was doing. In addition it must also be noted the reluctance of some scholars to conduct in-depth searches because, even in our days, it is not just enough to press the keys on a computer to get more information on topics from a data base. For the above reasons and with Barry Arnold’s words(53 ) in mind, we think that the journal Statistica and its Editor Italo Scardovi has done meritorious work by publishing in 2001 the Italian text and the English translation of Gini’s main papers on inference and probability(54 ). The possibility of having the papers of prominent Italian scholars in the language which is universally used in science today is perceived as an important need by many non-Italian researchers(55 ). New proposals and interpretations and the enduring topicality of Gini’s writings(56 ) make this situation even more fluid from a scientific point of view. In fact as highlighted by Anna Treves (2001, 240) “the archival sources [...] are very poor of information” for certain aspects relating to the life of the scholars who had intercourse with Fascism and of course, the discovery of additional documents may shed new light on the past. (51 ) In reality these summaries published almost always Gini’s work. On this subject also see Treves (2001, 218 (footnote 32), 225 (footnote 54) and 227). (52 ) See Gini (1921, 1926b, 1930, 1931a,b, 1936, 1965). (53 ) Arnold argues that “any new result on income inequality is with probability close to 1 a reinvention of something previously published in Italian”, see Giorgi (2005, 310). (54 ) Gini (2001). (55 ) This need was confirmed when, following the publication in Metron (issue 1, April 2005) of the English translation of an article by Gini (1914) “On the measurement of the concentration and variability of characters”, I received, as Editor of the journal, many messages of support for the publication of other papers hard to find and of thanks from non-Italian scholars for having made available Gini’s pioneering work. Many of them, despite having carried out studies on the Gini concentration ratio, had until then never read the original article in which this inequality index was proposed. (56 ) The topicality concerns also other lesser-known works. Di Bacco et al. (2006) have shown, by using the Bayesian paradigm and information from sources of historical demography and from an archaeological-anthropological sample, the validity of Gini’s (1957b) assumptions that the disappearance of the Viking colonies in Greenland was due to the imbalance of the genders.

20

GIOVANNI MARIA GIORGI

From the above mentioned biographical notes one can clearly infer Gini’s adherence to the general framework of Fascism, but he was a very special fascist! Although it is true that in 1925 he signed the “Fascist intellectuals manifesto” and in 1927 he wrote “The scientific basis of Fascism” but it is also true that he was too intelligent and “free spirit” to share all the ideas that the regime sought to impose. Thus, for example, his close ties with Mussolini during the presidency of Istat never made him subjugated from any point of view either scientific or human. Indeed during that period, he opposed the interference of the regime in his management of both staff and research. He prevented people without qualifications but with the fascist party card, to access positions for which he regarded them as not suitable. Instead, his rigid scientific and professional ethics led him to hire anti-fascist officials at Istat highest levels provided they were competent. Angelo Molinari was an outstanding example of this. As we mentioned in a previous section, he was called by Gini to be Istat general manager. In the purge processes that followed the end of Fascism, Molinari argued that “during the Gini management of Istat, any attempt to completely subjugate the Institute to Fascism [...] was impossible because of the strong and uncompromising opposition [of Gini] to any political interference in the affairs of the Institute”(57 ). Thus Molinari provided evidence, along with other eminent scholars and convinced antifascists, to drop the more serious charges against Gini(58 ). It should be noted that the strict adherence to scientific and professional ethics manifested through the defence of Istat from the external political interferences regarding the management of both staff and statistical data, does not mean that Gini was not in any way involved in some choices made by the Fascist state. For example the survey of large families (i.e. with more than 7 children) performed during the period 1928-31 and the forecasts of Italian population growth conducted during 1929-30 with Bruno de Finetti(59 ) in which the most favourable hypothesis (although not fully confirmed) approached “Mussolini’s aspiration to reach 60 million by 1960”(60 ). Therefore, the reasons of “science” and those of “merit” were the guidelines that characterized the academic and professional activities of Gini and the “reasons of science against those of the anti-Semitic and racial ideology”(61 ) led Gini to intervene in defence of the Cara¨ıtes in Lithuania and Poland. For (57 ) See Cassata (2006, 97). (58 ) Pr´evost (2009, 251) notes a “quite impressive” mobilization in favour of Gini of about a hundred people, including academics, Istat top managers and personalities from other institutions. Among the former was Paolo Fortunati (professor, Pietra’s collaborator and Senator of the Italian Communist party in 1948) who played a leading role by leveraging his anti-fascist and partisan militancy. (59 ) Gini and de Finetti (1931). (60 ) See Cassata (2006, 103-109) for further information on the large families survey and the forecast of Italian population. (61 ) Cassata (2006, 140).

Corrado Gini: the man and the scientist

21

the same reasons, after 8th September 1943, he prevented the intrusion of the Nazis and Fascists into (his) Faculty of Statistics(62 ). Historians are also investigating the role of Gini in the regime’s natality policy. Cassata (2006, 14) confirms the hypothesis of Gini’s marginal role in the early Fascist period as archival documents and Gini’s very frequent correspondence with Mussolini seem to confirm. However, Gini did have a supporting technical role as President of Istat. And it seems that he had provided Mussolini with data after his direct requests during the so-called speech of the Ascension(63 ) made in front of the Italian Parliament on 26th May 1927. In this speech Mussolini outlined the Fascist policy to stimulate the birth rate. However, Gini did not share and opposed(64 ) the choices made for families like those based on the marriage loan used in Nazi Germany and instituted by Mussolini in 1937. The behaviour of Gini, both as a scientist and as a public man, was quite varied, contradictory and not easily framed in the fascist scheme of that time. This makes us wonder what kind of a Fascist he really was! Above all he was an intellectual with a cultural background and mentality who could not be regarded as a real Fascist. He assessed “events and political decisions on the basis of their correspondence to his own schemes and theories”(65 ). He said in his self-defence at the purge process: “I am not a politician but a scientist, and as such, I consider my task is to observe how the events take place and to find the causes: even when these facts are in the field of politics I do not elaborate them as a politician, but as a scientist”(66 ). In reality the most relevant results of his scientific work and ideas - including the cyclical theory of nations and the foundations of neo-organicism(67 )(62 ) Cassata (2006, 127). (63 ) See Treves (2001, 212-217). (64 ) See Cassata (2006, 133). (65 ) Treves (2001, 227). (66 ) See Cassata (2006, 129). (67 ) The research program known as Gini neo-organicism was outlined synthetically in 1909 in his monograph “Il diverso accrescimento delle classi sociali e la concentrazione della ricchezza” and presented in a more extensive way in “I fattori demografici dell’evoluzione delle nazioni”(1912) and completed in the fourth edition (1935) of “Prime linee di patologia economica” which was further revised and enlarged in a fifth edition (1952) entitled “Patologia economica”. Neo-organicism identified the social organism with the living one characterized by an evolution from birth to extinction. In this context Gini saw nations as biological organisms. Neo-organicism was not exempt from criticism even from prominent members of the regime (e.g., De Stefani). For a critical vision of the neo-organicistic theory of society that Gini also supported by statistical arguments see, for example, Favero (2004), Maiocchi (2004) and Cassata (2006). Instead Castellano (1965) and Federici (1967) provided a more nuanced, more benevolent and almost deferential interpretation of the Ginian neoorganicism. On this subject see also Gregor (1967). It should also be noted that Dagum (1987, 529) saw the Ginian neo-organicism as “a well-structured anticipation of Wiener’s cybernetics, von Bertlanffy’s general system theory and modern disequilibrium economics”. Even Cassata (2006, 213) attributes to Gini and particularly to Castellano, the intuition and the development of “links between

22

GIOVANNI MARIA GIORGI

had already been outlined from the beginning of his academic career and well before the advent of Fascism. So he only shared with it that which was consistent with his own beliefs. In other words, we can say, paraphrasing Anna Treves (2001, 228), that he was a fascist only insofar as “he could consider Fascism as Ginian”, i.e. he was fascist when Fascism shared his ideas. He was, like most of the Italians of that time, a strong advocate of nationalism and from a political-culture point of view his thinking was of the renaissance type with Italy as cornerstone. One example of his personal way of being fascist can also be found in his “natural” adhesion to corporatism that was consistent with his vision of society as an economic organism(68 ). As Gini himself wrote in 1935 in the preface of the fourth edition of his book Patologia Economica(69 ), this vision had matured from his observations of the complex economic phenomena that characterized the World War I and the post-war period. Gini had observed these phenomena from privileged positions as head of the statistical office of the Ministry of War, as a member of the Committee on post-war problems, as director of the survey on raw materials carried out by the League of Nations (1921) and finally as an expert in the question of raw materials at the International Economic Conference of Genoa (1922). Like all brilliant minds, Gini had many interests. He was an economist, a statistician, a demographer, a sociologist, an anthropologist, i.e. a multi-faceted scientist. He was guided by an innate sense of scientific ethics which always enabled him to put scientific reasoning against that of politics. This way of being, we might say, his ethics was manifested on a daily basis both by refusing to bend himself - as President of Istat and as Dean of the Faculty of Statistics - to the favoritism “claimed” by the regime and by safeguarding the truth of statistical data through a rigorous use of statistical methods to avoid distortions. In Gini’s opinion, the use of suitable tools for the study of the phenomena under investigation and the verification of results were the “key”. Political ideology, when present, could be added but never imposed on the scientific structure and results of surveys. That is, he believed in the independence of the statistical procedure and its results which must not be bent to the needs of political power. We have already given an example of this in Section 3 which highlights Gini’s support to Molinari especially for the dissemination and publication of Census results of 1921 which went against the expectations of the regime. Another case of this was the 1931 census when neo-organicism and systems theory that Robert Merton had already noticed in the mid-thirties in his review of Gini’s Patologia economica”. For a synthetic analysis and some interesting observations on Castellano’s neo-organicism see also Sgritta (2001). (68 ) Corporatism was seen by Mussolini as an alternative to capitalism whose crisis was explained by Gini in terms of “economic pathology”. He also noted that the capitalism transformation had already started since World War I. For the interested reader a discussion of various aspects of the Ginian theory of society as an economic organism can be found in Gini (1952) and also in Fortunati (1942). (69 ) Gini (1952, xxvii-xxxi).

Corrado Gini: the man and the scientist

23

Gini found that some census data had been altered to show that Catania and Naples populations were greater than they actually were. As a result Gini decided to repeat the census in those two cities(70 ). In an attempt to better delineate the complex personality of Gini, we have primarily focused on human aspects and the need to frame him from a perspective less aseptic than one might infer from the cold archival material has led us to take account of memories emerged from many conversations, partially reported in two previous interviews (Giorgi, 1996, 2002), we had with those who knew him closely. In addition, historical consciousness reminds us to include the period in which Gini lived with particular reference to the Fascist era. In conclusion we can say that Corrado Gini - the man and the statistician who dominated the Italian statistical scene for half a century - is far from a forgotten or irrelevant figure. Further research, especially that based on new materials from his archives and private library, will continue to reveal more about this extraordinary and controversial personality. And we can say with certainty that Corrado Gini will continue to be a fascinating figure to study. REFERENCES Baffigi, A. and Magnani, M. (2009) Giorgio Mortara, “Le leggi antiebraiche del 1938, le societ`a scientifiche e la scuola in Italia”, Roma, Accademia Nazionale delle Scienze detta dei XL, Scritti e documenti, 42, 237–253. ´ S. and Pales, ´ Bajak, Z. (2009) Computer Aided Solution of the Invariance Equation for TwoVariable Gini Means, Computers and Mathematics with Applications, 58 (2), 334–340. Benedetti, C. (1984) A proposito del centenario della nascita di Corrado Gini, Metron, 42 (1-2), 3–19. Bertaux, S.(1999) D´emographie, statistique et fascisme. Corrado Gini et l’Istat, entre science et ideologie (1926-1932), Roma moderna e contemporanea, 7 (3), 571–598. Biggeri, L., Ferrari, G. and Lemmi, A. (1987) The Contribution of Italian Statisticians to the Theory of Price Index Numbers, In: Italian Contribution to the Methodology of Statistics (A. Naddeo (ed.)), Societ`a Italiana di Statistica, Padova, Cleup, 563–581. Boldrini, M. (1942) Recenti trattati italiani nelle discipline statistiche, Statistica, 2 (4), 335–348. Bonferroni, C. E. (1940) Elementi di statistica generale, Torino, Gili. Bortkiewicz, L. (1931a) Die Disparit¨atsmasse der Einkommensstatistik, Bulletin de l’Institut International de Statistique, 19`eme Session (Tokio,1930), tome 25, 3`eme livraison, 2`eme partie, 189–298. Bortkiewicz, L. (1931b) Erwiderung, Bulletin de l’Institut International de Statistique, 19`eme Session (Tokio,1930), tome 25, 3`eme livraison, 2`eme partie, 311–316. Cassata, F. (2004) Cronaca di un’epurazione mancata, Popolazione e Storia, 5 (2), 89–119. Cassata, F. (2006) Il fascismo razionale. Corrado Gini fra scienza e politica, Roma, Carocci. (70 ) On this subject see Cassata (2006, 99).

24

GIOVANNI MARIA GIORGI

Cassese, S. (2010) Lo Stato fascista, Bologna, il Mulino. Castellano, V. (1965) Corrado Gini: A Memoir, Metron, 24 (1-4), 3–84. Castellano, V. (1996) I sessantanni della Facolt`a di Scienze Statistiche: La storia, Atti della Manifestazione, Roma, Universit`a “La Sapienza”, 10–13. Cocchi, D. and Favero, G. (2009) Gli statistici e la “questione della razza”. Le leggi antiebraiche del 1938, le societ`a scientifiche e la scuola in Italia, Roma, Accademia Nazionale delle Scienze detta dei XL, Scritti e documenti, 42, 207–235. Conti, P. L. and Giorgi, G. M. (2001) Distribution-Free Estimation of the Gini Inequality Index: The Kernel Method Approach, Statistica, 61 (1), 5–14. Cuthbert, J. R.(2003) On the Variance/Covariance Structure of the Log Fisher Index, and Implications for Aggregation Techniques, Review of Income and Wealth, 49 (1), 69–88. ´ Czinder, P. and Pales, Z. (2005) Local Monotonicity Properties of Two-Variable Gini Means and the Comparison Theorem Revisited, J. Mathematical Analysis and Applications, 301 (2), 427–438. Dagum, C. (1987) Gini Corrado (1884-1965), In: The New Palgrave: A Dictionary of Economics (J. Eatwell, M. Milgate and P. Newmann (eds.)), London, Macmillan Press Limited, 2, 529. Dalla Zuanna, G. (ed.) (2004) Numeri e potere: Statistica e demografia nella cultura italiana tra le due guerre, Napoli, L’ancora del Mediterraneo. D’Autilia, M. L. (1992) L’amministrazione della statistica nell’Italia fascista. Il caso dell’agricoltura, Roma, Gangemi Editore. De Sandre, P. (2004) Demografia politica e politiche di popolazione nella cultura italiana del ventennio, In: Numeri e potere. Statistica e demografia nella cultura italiana tra le due guerre (G. Dalla Zuanna (ed.)), Napoli, L’ancora del Mediterraneo, 43–68. Di Bacco, M., Del Panta, L., Frederic, P., D’Amore, G. and Lynnerup, N. (2006) The Effect of an Unbalanced Demographic Structure on Marriage and Fertility Patterns in Isolated Populations: the Case of Norse Settlements in Greenland, Genus, 62 (1), 97–119. Donati, A. (1996) I sessantanni della Facolt`a di scienze statistiche: La storia, Atti della Manifestazione, Roma, Universit`a “La Sapienza”, 9–10. Farnsworth D. and Orr, R. (1986) Gini Means, American Mathematical Monthly, 93 (8), 603–607. Favero, G. (2004) Corrado Gini and Italian Statistics under Fascism, Il Pensiero Economico Italiano, 12 (1), 45–59. Favero, G. and Trivellato, U. (2009) La statistica nell’Universit`a di Padova prima della nascita della Facolt`a, In: Una lunga storia giovane. La Statistica a Padova fra 200 anni di storia e 40 anni della Facolt`a, Padova, Cleup, 13–80. Federici, N. (1960) L’opera di Corrado Gini nell’ambito delle scienze sociali, In: Studi in Onore di Corrado Gini, Roma, Facolt`a di Scienze Statistiche dell’Universit`a di Roma, 2 , 3–34. Federici, N. (1967) Riflessioni sulla teoria ciclica dell’evoluzione delle popolazioni e sulla concezione neo-organicista della societ`a, Atti del simposio internazionale su “La statistica come metodologia delle scienze sociali”, Roma, 13-15 Marzo 1966, Istituto di Statistica e Ricerca Sociale “C. Gini” dell’Universit`a di Roma, 3, 51–65. Federici, N. (2000) Gini Corrado, Dizionario biografico degli italiani, Roma, Istituto della Enciclopedia Italiana, 55, 18–21. Forcina, A. (1982) Gini’s Contribution to the Theory of Inference, International Statistical Review, 50 (1), 65–70. Fortunati, P. (1942) Fisiologia e patologia economica, Statistica, 2 (4), 348–365.

Corrado Gini: the man and the scientist

25

Frick, J. R., Goebel, J., Schechtman, E., Wagner, G. G. and Yitzhaki, S. (2006) Using Analysis of Gini (ANOGI) for Detecting Whether Two Sub-Samples Represent the Same Universe, Sociological Methods & Research, 34 (4), 427–468. Frish, R. (1936) Annual Survey of General Economic Theory: The Problem of Index Numbers, Econometrica, 4 (1), 1–38. Gagliardi, A. (2010) Il corporativismo fascista, Bari, Laterza. Georgescu-Roegen, N. (1966) Further Thoughts on Corrado Gini’s delusioni dell’econometria, Metron, 25 (1-4), 265–279. Gini, C. (1908) Il sesso dal punto di vista statistico, Napoli, Sandrom. Gini C. (1909) Il diverso accrescimento delle classi sociali e la concentrazione della ricchezza, Giornale degli Economisti, series 2, 38 (1), 27–83. Gini, C. (1910) Prezzi e consumi, Giornale degli Economisti e Rivista di Statistica, series 3, 40 (1, 2), 99–114 and 235–249. Gini C. (1911) Considerazioni sulle probabilit`a a posteriori ed applicazioni al rapporto dei sessi nelle nascite umane, In: Studi Economico-Giuridici della Facolt`a di Giurisprudenza della Regia Universit`a di Cagliari, reprinted in Metron, 1949, 15 (1-4), 133–172. Gini, C. (1912) I fattori demografici dell’evoluzione delle nazioni, Torino, Bocca. Gini, C. (1914) Sulla misura della concentrazione e della variabilit`a dei caratteri, Atti del Reale Istituto Veneto di Scienze, Lettere ed Arti, 73 , part 2, 1203–1248; English translation in Metron, 2005, 63 (1), 3–38. Gini, C. (1921) Measurement of Inequality of Incomes, Economic J., 31 (121), 124–126. Gini, C. (1924) Quelques consid´erations au sujet de la construction des nombres indices des prix et des questions analogues, Metron, 4 (1), 3–162. Gini, C. (1926a) Le basi teoriche della politica economica, Economia, 8 (2), 113–143; English translation in Journal of Political Economy, 1929, 37 (6), 633–660. Gini, C. (1926b) The Contributions of Italy to Modern Statistical Methods, J. Royal Statistical Society, 89 (4), 703–724. Gini, C. (1927) The Scientific Basis of Fascism, Political Science Quarterly, 42 (1), 99–115. Gini, C. (1930) Present Condition and Future Progress of Statistics, J. American Statistical Association, 25 (171), 295–304. Gini, C. (1931a) Observations a` la communication du Prof. L. von Bortkiewicz, Bulletin de l’Institut International de Statistique, 19`eme Session (Tokio, 1930), tome 25, 3`eme livraison, 2`eme partie, 299–306. Gini, C. (1931b) Note, Bulletin de l’Institut International de Statistique, 19`eme Session (Tokio, 1930), tome 25, 3`eme livraison, 2`eme partie, 317–320K. Gini, C. (1936) On the Measure of Concentration with Special Reference to Income and Wealth, Cowles Commission Research Conference on Economics and Statistics, July 6-August 8, Colorado College Publications, General Series, n. 208, 73–80. Gini, C. (1938) Di una formula comprensiva delle medie, Metron, 13 (2), 3–22. Gini, C. (1939) I pericoli della statistica, Supplemento statistico ai nuovi problemi di politica, storia ed economia, series 2, 5 (2-4), 1–44. Gini, C. (1943) I testi di significativit`a, Atti della VII Riunione Scientifica della Societ`a Italiana di Statistica, Roma, 27-30 giugno, 241–279. Gini, C. (1952) Patologia economica, (5th edition), Torino, Utet. Gini, C.(1956) Delusioni dell’econometria, Giornale degli economisti ed annali di economia, 15 (3-4), new series, 174–177.

26

GIOVANNI MARIA GIORGI

Gini, C. (1957a) Le medie, Torino, Utet. Gini, C. (1957b) Sulla scomparsa delle colonie normanne in Groenlandia, Corso di sociologia, Roma, Edizioni Ricerche, chap. 9, 682–715. Gini, C. (1958) Ricchezza e reddito, Torino, Utet. Gini, C. (1965) On the Characteristics of Italian Statistics, J. Royal Statistical Society, series A, 128 (1), 89–109. Gini, C. (2001) Statistica e Induzione, Biblioteca di Statistica, Bologna, Clueb. Gini, C. and de Finetti, B. (1931) Calcoli sullo sviluppo futuro della popolazione italiana, Annali di statistica, Roma, Istat, series 6, 10. Giorgi, G. M. (1984) Alcune considerazioni teoriche su di un vecchio ma pur sempre attuale indice: il rapporto di concentrazione, Metron, 42 (3-4), 25–40. Giorgi, G. M. (1990) Bibliographic Portrait of the Gini Concentration Ratio, Metron, 48 (1-4), 183–221. Giorgi, G. M. (1992) Il Rapporto di concentrazione di Gini: genesi, evoluzione ed una bibliografia commentata, Siena, Editrice Ticci. Giorgi, G. M. (1993) A Fresh Look at the Topical interest of the Gini Concentration Ratio, Metron, 51 (1-2), 83–98. Giorgi, G. M. (1996) Encounters with the Italian Statistical School: A Conversation with Carlo Benedetti, Metron, 54 (3-4), 3–23. Giorgi, G. M. (1998) Concentration Index, Bonferroni, In: Encyclopedia of Statistical Sciences (S. Kotz, C.B. Read and D.L. Banks (Eds.)), New York, Wiley & Sons, update, 2, 141–146. Giorgi, G. M. (1999) Income Inequality Measurement: The Statistical Approach, In: Handbook on Income Inequality Measurement (J. Silber (ed.)), Boston, Kluwer Academic Publishers, 245–260. Giorgi, G. M. (2001) Corrado Gini, In: Statisticians of the Centuries (C.C. Heyde and E. Seneta (eds.)), New York, Springer-Verlag, 364–368. Giorgi, G. M. (2002) Intervista a Italo Scardovi, Statistica & Societ`a, 1 (1), 15–23. Giorgi, G. M. (2005) Gini’s Scientific Work: An Evergreen, Metron, 63 (3), 299–315. Giorgi, G. M. (2010) Some Considerations on Pr´evost Book: “A Total Science: Statistics in Liberal and Fascist Italy”, Metron, 68 (1), 105–110. Giorgi, G. M. (2011) The Gini Inequality Index Decomposition: An Evolutionary Study, In: The Measurement of Individual Well Being and Group Inequalities: Essays in Memory of Z. B. Berebi (J. Deutsch and J. Silber (eds.)), Oxon, Routledge, 185–218. Giorgi, G. M. and Nadarajah, S. (2010) Bonferroni and Gini Indices for Various Parametric Families of Distributions, Metron, 68 (1), 23–46. Giorgi, G. M. and Pallini, A. (1986) Di talune soglie inferiori e superiori del rapporto di concentrazione, Metron, 44 (1-4), 377–390. Giorgi, G. M. and Pallini, A. (1987a) About a General Method for the Lower and Upper Distribution-free Bounds on Gini’s Concentration Ratio from Grouped Data, Statistica, 47 (2), 171–184. Giorgi, G. M. and Pallini, A. (1987b) On the Speed of Convergence to the Normal Distribution of Some Inequality Indices, Metron, 45 (3-4), 137–149. Giorgi, G. M. and Pallini, A. (1989) Upper Bounds on Gini Index: a Measure for Geometric Comparison, Economic Notes, 19, 282–286. Giorgi, G. M. and Pallini, A. (1990) Inequality Indices: Theoretical and Empirical Aspects of Their Asymptotic Behaviour, Statistical Papers, 31 (1), 65–76.

Corrado Gini: the man and the scientist

27

Gregor, A. J. (1967) Some Aspects of the Gini Scientific Work of Corrado Gini, Atti del simposio internazionale su “La statistica come metodologia delle scienze sociali”, Roma, 13-15 Marzo 1966, Istituto di Statistica e Ricerca Sociale “C. Gini”, Universit`a di Roma, 3, 66–77. Herzel, A. and Leti, G. (1997) Italian Contributions to Statistical Inference, Metron, 35 (1-2), 3–48. Kendall, M. G. and Buckland, W. R. (eds.) (1982) Dictionary of Statistical Terms, London, Longman Group, 4th edition. Kleiber, C. and Kotz, S. (2003) Statistical Size Distributions in Economics and Actuarial Sciences, Hoboken (NJ), Wiley. Leti, G. (1996) L’Istat e il Consiglio superiore di statistica dal 1926 al 1945, Annali di statistica, Roma, Istat, series 10, 8. Maiocchi, R. (2003) Gli scienziati del duce, Roma, Carocci. Maiocchi, R. (2004) La demografia italiana e il fascismo, In: Numeri e potere. Statistica e demografia nella cultura italiana tra le due guerre (G. Dalla Zuanna (ed.)), Napoli, L’ancora del Mediterraneo, 15–42. Mieli, P. (2001) Cos`ı Fermi scopr`ı la natura vessatoria del fascismo; In: Lettere al Corriere, Milano, Corriere della Sera - 2 Ottobre 2001, 41. Misiani, S. (2007) I numeri e la politica. Statistica, programmazione e mezzogiorno nell’impegno di Alessandro Molinari, Bologna, il Mulino. Molinari, A. (1932) Un’indagine sulle motivazioni della diminuzione delle nascite, Metron, 10 (1-2), 349–391. Montanari, A. and Monari P. (2008) Gini’s Ideas: New Perspectives for Modern Multivariate Statistical Analysis, Statistica, 68 (3-4), 239–254. Montevecchi, L. (2002) Una recente acquisizione dell’Archivio Centrale dello Stato: L’archivio e la biblioteca di Corrado Gini, Popolazione e Storia, 3 (2), 155–157. ´ Neuman, E. and Sandor, J. (2003) Inequalities Involving Stolarsky and Gini Means, Mathematica Pannonica, 14 (1), 29–44. Olkin, I. and Yitzhaki, S. (1992) Gini Regression Analysis, International Statistical Review, 60 (2), 185–196. Pedroni, F. (1960) Configurazione di uno schema teorico per la determinazione del reddito nazionale secondo i concetti di Corrado Gini; In: Studi in Onore di Corrado Gini, Roma, Facolt`a di Scienze Statistiche dell’Universit`a di Roma, 1, 515–534. Perfetti F. (2010) Lo stato fascista. Le basi sindacali e corporative, Firenze, Le Lettere. Piccinato, L. (2011) Gini’s Criticisms to the Theory of Inference: a Missed Opportunity, Metron, 69, this issue. Pietra, G. (1931a) Observations a` la communication du Prof. L. von Bortkiewicz, Bulletin de l’Institut International de Statistique, 19´eme Session (Tokio,1930), tome 25, 3`eme livraison, 2`eme partie, 310. Pietra, G. (1931b) Note, Bulletin de l’Institut International de Statistique, 19´eme Session (Tokio, 1930), tome 25, 3`eme livraison, 2`eme partie, 320L-320M. Pr´evost, J. G. (2001) Une pathologie politique: Corrado Gini et la critique de la d´emocratie lib´erale, Revue franc¸aise d’histoire des id´ees politiques, 13 (1), 105–128. Pr´evost, J. G. (2009) A Total Science. Statistics in Liberal and Fascist Italy, Montreal, McGillQueen’s University Press. Romano, S. (2006) 1931: i professori giurano fedelt`a al fascismo; In: Lettere al Corriere, Milano, Corriere della Sera - 14 Febbraio 2006, 39. Roy, R. (1966) R´eflexions inspir´ees par quelques recherches de Corrado Gini, Metron, 25 (1-4), 280–290.

28

GIOVANNI MARIA GIORGI

Savorgnan, F. (1931) Observations a` la communication du Prof. L. von Bortkiewicz, Bulletin de l’Institut International de Statistique, 19´eme Session (Tokio,1930), tome 25, 3`eme livraison, 2`eme partie, 307–309. Schechtman, E. and Yitzhaki, S. (1987) A Measure of Association Based on Gini’s Mean Difference, Communications in Statistics-Theory and Methods, 16 (1), 207–231. Schechtman, E. and Yitzhaki, S. (1999) On the Proper Bounds of the Gini Correlation, Economics Letters, 63 (2), 133–138. Schechtman, E. and Yitzhaki, S. (2003) A Family of Correlation Coefficients Based on Extended Gini, J. Economic Inequality, 1 (1), 129–146. Schechtman, E., Yitzhaki, S. and Pudalov, T. (2011) Gini’s Multiple Regressions: Two Approaches and Their Iteraction, Metron, 69, this issue. Sedita, G. (2010) Gli intellettuali di Mussolini. La cultura finanziata dal fascismo, Firenze, Le Lettere. Sen, A. (1999) Foreword; In: Handbook on Income Inequality Measurement (J. Silber (ed.)), Boston, Kluwer Academic Publishers, xvii-xxvi. Sgritta, G. B. (2001) La societ`a come sistema, In: Vittorio Castellano statistico e sociologo, Roma, Facolt`a di Scienze Statistiche dell’Universit`a di Roma, 71–88. Shi, H-N., Jiang, Y-M. and Jiang, WW. D. (2009) Schur-Convexity and Schur-Geometrically Concavity of Gini Means, Computers and Mathematics with Applications, 57 (2), 266–274. Treves, A. (2001) Le nascite e la politica nell’Italia del novecento, Milano, Edizioni Universitarie di Lettere, Economia e Diritto. Treves, A. (2003) Demografi, fascismo e politica delle nascite. Nodi problematici e prospettive di ricerca, Popolazione e Storia, 4 (1), 183–194. Trivellato, U. (2004) Al crocevia fra scienza, ideologia e regime: uno sguardo allo sfondo e ad alcuni statistici e demografi eminenti, In: Numeri e potere. Statistica e demografia nella cultura italiana tra le due guerre (G. Dalla Zuanna (ed.)), Napoli, L’ancora del Mediterraneo, 69–86. Ventura, A. (1996) Le leggi razziali all’Universit`a di Padova, In: L’Universit`a dalle leggi razziali alla resistenza (A. Ventura (ed.)), Padova, Cleup, 131–204. Yitzhaki, S. (1991) Calculating Jackknife Variance Estimators for Parameters of the Gini Method, J. Business & Economic Statistics, 9 (2), 235–239. Yitzhaki, S. and Olkin, I. (1991) Concentration Indices and Concentration Curves, In: Stochastic Orders and Decision under Risk (Mosler K. and Scarsini M. (eds.), IMS Lecture NotesMonograph Series, Hayward (CA), 19, 380–392. Yule, G. U. and Kendall, M. G. (1940) Introduction to the Theory of Statistics, London, Griffin & Co.

GIOVANNI MARIA GIORGI Dipartimento di Scienze Statistiche “Sapienza” Universit`a di Roma Piazzale Aldo Moro, 5 00185 Roma (Italy) [email protected]