Inequality Within the Family: Cases of Selective Parents in Post-War Hong Kong

by Yi-Lee Wong
Tohoku University

Sociological Research Online 12(5)15

Received: 12 Mar 2007     Accepted: 2 Sep 2007    Published: 30 Sep 2007


While members of the same family are assumed to share similar mobility chances, this paper seeks to answer the following puzzle: why do only some children of the same family attain a level of education considered to be socially desirable whereas their siblings do not? The essence of an answer lies in the fact that the same parents could play rather dissimilar roles in the education of their different children. Using part of qualitative data collected in Hong Kong between 1996 and 1997, this paper focuses on what selective parents did for their children's education. The data illustrated that in deciding what they would and could do for each of their children's education, parents responded to their children's academic ability, resource availability, and ideology. The educational attainments of children of the same family could be very diverse not merely because of children's different academic performances but because of the deliberate decisions of their parents in formulating strategies for basic survival or for advancement. The same parents could be seen as enhancing the education of their sons and/or younger children at the expense of the education of their daughters and/or elder children. This suggests that mobility is of an interdependent nature and, in turn, leads me to argue that the mobility of members of the same family should be considered together and not in isolation, and to support the stance that the family, rather than an individual, should be the unit of analysis in mobility studies.

Keywords: Birth Order; Education; Family; Gender; Selective Parents; Social Mobility


1.1 As teachers teaching courses on social stratification, we all introduce our students to the idea that people of the same class share similar life chances. In particular, mobility studies have shown that people of the same class share similar mobility chances. In their description and/or explanation of mobility patterns, mobility researchers make certain assumptions about mobility and the family. Mobility is basically seen as an individual pursuit in that people of different class origins are competing with each other for an advantaged level of education, a prestigious job, and a socially desirable class position. Their parents are taken as one of many independent variables in affecting their final class destination playing the role of providing them with socialisation and/or resources. So, in explaining class differentials in educational attainment, sociologists have been debating over whether this class gap reflects variations in upbringing, culture, or orientation towards education across classes (e.g. Bourdieu and Passeron 1977) or whether it reflects differences in resources between classes (e.g. Boudon 1974). However, these two, competing, explanations both fail to explain differences within the family. What is puzzling is this: coming from the same family, children presumably have a similar kind of upbringing and adopt a similar orientation towards education, or supposedly they receive a similar amount of resources from the same parents; but, why do only some children attain a level of education considered to be socially advantaged whereas their siblings do not?

1.2 Our intuition may lead us to focus on individual children’s academic ability: siblings differ in biological make-up and even twins may also differ in genetic endowment. Some psychologists link this to sibling rivalry and child development comparing the environment in which children of different birth orders grow up. It is argued that in most families it is usually first-born children that have the greatest success in adult life because they are more advantaged than their younger siblings in terms of biological and intellectual developments. By nature, first-born children are born to be physically stronger and more intelligent than their younger siblings (e.g. Zajonc and Markus 1975). As the likelihood of birth defects increases with mother’s age at birth, first-born children are less likely to be biologically defective in any way and thus develop better biologically and intellectually vis-à-vis their younger siblings. By nurture, first-born children grow up in an environment more conducive to their intellectual development than their younger siblings. They grow up in an adult-oriented environment and have at least one more year of exclusive one-on-one time with their parents than their younger siblings (e.g. Leman and Leman 1987). Besides, first-born children usually take up the task of teaching their younger siblings; but they actually learn more from teaching their younger siblings than their younger siblings learn from first-born children’s teaching (cf. Sulloway 1996). Whether birth order has an effect on the development of individuals and of their personalities remains controversial (e.g. Steelman et al. 2002). But, these psychological studies do not seem to pay much attention to wider social forces. It is well-documented that the effects of academic ability on one’s educational attainment are not independent of the impact of social forces and that academic ability alone does not explain sufficiently differences in educational attainment (cf. Savage and Egerton 1997).

1.3 Perhaps what matters is not merely children’s academic ability but how parents respond to it. Three types of parents distinguished by some economists are of relevance here (e.g. Mulligan 1997). First, parents are neutral and thus invest in each child’s education in a similar fashion and pass on to each child a similar amount of financial assets. Second, parents favour academically-most-capable children and thus invest more in their education; this ultimately reinforces siblings’ differences in genetic endowment and academic ability, manifesting itself through siblings’ differentials in educational attainment and earning ability. Third, parents compensate academically-least-capable children and thus pass on to them more financial assets; this compensates siblings’ differences in educational attainment and thus earning ability, manifesting itself in similarity in siblings’ possession of wealth (i.e. the sum of earning and financial assets). This suggests that parents have strategies for their children.

1.4 Whether parents could implement their strategies is contingent on the availability of resources. The variables of the number and the spacing of children and the concept of the family-life-cycle have been considered in relation to resource availability to explain differences in siblings’ educational attainment: resources available at various stages of the family-life-cycle constrain what parents can do for the education of their children of different birth orders (e.g. Van Eijck and De Graaf 1995; Guo and Van Wey 1999). Nevertheless, resource availability does not explain sufficiently what parents would do for the education of their children of different birth orders or for that of their sons and daughters. In addition to referring to ideology that parents hold about proper gender roles, various versions of the notion of ‘patriarchy’ have been advanced (e.g. Walby 1990). But then, resource availability and ideology seem to be treated as two additional variables and birth order and gender as two separate domains of inequality within the family in that the domain of birth order is explained only by resource availability whereas that of gender is explained by ideology alone. It is possible that parents discriminate against their children of different birth orders for ideological reasons. As an example, two practices have been commonly found in Europe since the sixteenth century: primogeniture and ultimogeniture, the right of inheritance belonging to the eldest son and to the youngest son respectively (Burguiere et al. 1996). Similarly, the sexist practices of parents could also be explained in terms of resource availability: given the existence of a wage gap in favour of men in the labour market and assuming that their children will return what parents do for the children’s education in monetary terms, parents may find it more profitable to invest more in the education of their sons than that of their daughters. Besides, the possibility that there could be interactive effects of resource availability and ideology seems ignored.

1.5 Despite having rather different focus in their explanations, these studies all point to the fact that siblings may not share the same advantages that their class or family confers. This suggests that, first, mobility chances of children of the same family are not necessarily identical; second, the role of the same parents in promoting the educational success of their different children could very well be dissimilar; and third, mobility is not purely an individual pursuit but could involve negotiations with one’s parents and/or siblings. This points to the need for paying attention to family dynamics and examining closely the practices of selective parents in treating each of their children. While succeeding in presenting general patterns of parental treatment for their children, quantitative data do not provide ready access to contextualised processes to illuminate the treatment of the same parents for their different children. Drawing on qualitative data on selective parents, I shall examine what they do in promoting the educational success of their different children. In what follows, I shall first describe the design of this research project and its context – post-war Hong Kong. I shall then move on to a discussion of my results concerning these selective parents. Finally, I shall answer the puzzle posed above and conclude this paper by supporting the stance that the family, rather than an individual, should be taken as the unit of analysis in mobility studies.

Research design

2.1 As an industrial-capitalist society exhibiting similar mobility patterns to its counterparts in the West, Hong Kong could be used as an example for illustration (cf. Chan 1994). In order to understand differentials in siblings’ educational attainment, I suggest that it be crucial to examine closely the role of the same parents in the education of their children. And, in order to see how the same parents could play dissimilar roles for their different children, we should look into the practices of selective parents. Without longitudinal data, a retrospective rather than a prospective design seems slightly more desirable. That is, rather than examining what parents do for their children now and predicting what level of education the children will achieve later on, I seek to take children’s educational attainments as an outcome and to explain it by tracing what their parents did for them in the past.

2.2 Data of this paper are drawn from part of a qualitative study on mobility strategies of teachers and managers – two occupational groups of Goldthorpe’s (Erikson and Goldthorpe 1992:39-42) service-class professional and managerial employees – in contemporary Hong Kong.[1] One major goal of this study is to examine how and why some particular people succeed in getting a middle-class position, a socially desirable position in many industrial-capitalist societies. Teachers and managers are selected[2] because they are two rather broad categories covering a range of middle-class occupations: the category of teachers includes primary school teachers as well as university lecturers while the category of managers includes not only small firm managers but also international corporation chief executive officers. Because of this relatively wide coverage, their experiences might represent the general experience of the middle class. Having set the selection criteria, I began to recruit targeted respondents by sending recruitment letters to a number of big firms, banks, organisation, secondary schools, tertiary institutions, and universities; only three respondents were recruited. Meanwhile, I exhausted my personal networks and exploited social networks of my relatives, friends, and acquaintances as much as I could for recruitment; most respondents were recruited in this way. I then asked the interviewed respondents to introduce me to further qualified respondents; but this snowballing process was not very successful: only four more respondents were recruited. In short, through both formal and informal channels, eighty-nine respondents – teachers and managers and their respective spouses – were recruited. For present purposes, this paper reports only data for thirty respondents who reported that their parents were selective in giving support to their education and to that of their siblings.

2.3 Data were collected through in-depth interviews between September 1996 and August 1997. The interviews were taped and ranged from forty-five minutes to two-and-half hours (the average being an hour); they were then transcribed and translated from Cantonese[3] to English. One main theme of this study was the role of parents in their children’s social mobility (i.e. their children’s education and career); and I sought to examine it from two perspectives. First, taking respondents as children, I compared what their parents did for the education and career of the respondents with what they did for those of the respondents’ siblings. Second, taking respondents as parents, on the one hand, I examined what teacher-respondents and manager-respondents did for their children’s education to see how they differed, while on the other hand, I compared what parents of the respondents had done for their education with what the respondents did for the education of their own children. Data were analysed thematically: the focus was on parents’ aspirations for their children, resources used, rationales behind their actions, and on children’s aspirations and reactions to their parents’ actions. The data reported below are from the detailed descriptions of respondents about their educational history as well as that of their siblings, and about what their parents did for their education and for that of their siblings. Out of thirty respondents reported in this paper, eighteen were male and twelve female. On average they were aged about forty-six (in 1997), ranging from thirty-five to fifty-nine years of age. Consistent with the data for that period, respondents had on average four to five siblings, ranging from three to twelve (Hong Kong Census and Statistics Department).

2.4 Before going on to discuss my findings and their implications, I would like to draw readers’ attention to two concerns about the data reported in this paper. The first is about the sample. It is a small non-random sample. Referring to specific selection criteria, I recruited so-called successful people – those having a socially desirable teaching or managerial job or those marrying people with this type of job. The primary concern of using these selection criteria was that I sought to explain how and why they became successful and to examine the role of their parents in their mobility successes; and as a by-product, data also allowed me to examine if their siblings were as successful as the respondents were and to explore mechanisms explaining sibling differences. Therefore readers are reminded that I do not seek to use data here to make statistical generalisations. The second concern is about the data. As with respondents in other retrospective studies, my respondents may not recall accurately what happened in the past, especially what their parents did for their siblings, and they may even rationalise their experiences by choosing apposite episodes to fit in with their own story lines (Elder 1981). It would have been ideal if I had interviewed the parents and siblings of my respondents so as to illustrate the complexity of family dynamics. However, when this study was carried out, most parents of the respondents were either dead or living in other countries, while some siblings of the respondents were living abroad. Because of these practical difficulties, I did not interview any parents or siblings of my respondents. Respondents were the only source of information reporting what parents did for their different children and how the children responded. Readers should be cautious because of unavoidable, conscious or unconscious, distortions in the respondents’ accounts. As I have no material from the perspectives of different family members, which definitely deserves more attention for future studies along this line of inquiry, readers should also bear in mind that what follows is an illustration of family dynamics only from the perspective of my respondents.

Post-war Hong Kong

3.1 A brief description of post-war Hong Kong would be useful for understanding the context against which parents acted in relation to the education of my respondents and that of their siblings. Before the Second World War Hong Kong was essentially an entrepot with its labour force engaged in trading, transport and communication, and maritime-related activities; industrialisation did not begin until the 1950s (cf. Endacott 1973). Most of my respondents grew up in Hong Kong of the 1950s and 1960s, when the great majority of people lived from hand to mouth. Basic education was not free or compulsory: even in the 1960s and 1970s, amongst the population aged fifteen or above, over a quarter had received no schooling and only about one-tenth had completed a secondary education (Hong Kong Census and Statistics Department); the acts of free, universal, and compulsory six-year and nine-year education were passed only in 1971 and 1978 respectively. Child labour was not uncommon; the law setting the minimum legal age at fourteen for all kinds of employment was passed in 1964 (Hong Kong Report for the Year 1964) and the law setting it at the age of fifteen came into effect only in 1980 (Hong Kong 1981: A Review of 1980).

3.2 Against the context of industrialisation, so-called traditional Chinese and modern Western values coexisted in post-war Hong Kong (Mitchell 1972; cf. Inglehart and Baker 2000). What is of particular relevance is a so-called traditional Chinese ideology of the family (cf. Baker 1979). It is commonly referred to in explaining the cultural specificity of Hong Kong vis-à-vis other western industrial-capitalist societies. According to this ideology, the family is continued through the male line: a son is to carry his family name, to inherit family property, and to get married in order to have at least one son to keep the family going. Generally speaking, both in the West and in Chinese societies, when a child is young and incapable, s/he is fed, clothed, housed, and educated by his/her parents. But in Chinese societies, when a son’s parents become old and weak, he is strongly expected to feed, clothe, shelter, and care for his parents, thus directly reciprocating their previous care of him. In the case of a daughter, the reciprocity is indirect in that it is her husband’s parents to whom she repays the care expended on her by her own parents. A son’s achievement is not only a personal success but an advancement of the whole family (constituted by all his ancestors in the past and future successors). What is implied in this ideology is that sons should advance through education and employment and daughters through marriage. While different family ideals exist in many societies, it is not uncommon to observe discrepancies between what is upheld as an ideal and what happens in reality. Describing this traditional Chinese ideology, I do not mean that all the Chinese in post-war Hong Kong subscribe to this ideology or practise it. Neither do I suggest that it is the only traditional Chinese ideology at work in post-war Hong Kong. But this particular ideology could somehow be seen as one of the most widespread common beliefs structuring Chinese parent-child interactions.

A typology of selective parents

4.1 According to what my respondents reported, three main groups of selective parents could be distinguished. The first is those who base their discriminatory treatment on birth order only: parents of twelve respondents were selective in treating their elder and younger children (EY parents thereafter).[4] The second is those who base their discriminatory treatment on gender only: parents of three respondents were selective in treating their sons and daughters (SD parents thereafter).[5] The third is those who base their discriminatory treatment much more on gender than birth order: parents of fifteen respondents were selective in treating their sons and daughters of different birth orders (SD-EY parents thereafter).[6] I shall turn to how these selective parents treated their different children and use the cases of employer SD-EY parents to highlight that parents could play a very direct role in the mobility of their different children.

EY parents

4.2 EY parents were selective not out of parental favouritism. And, contrary to the claim by psychologists that first-born children were most advantaged of all children, the data show that the elder, especially the eldest,[7] were more disadvantaged than their younger siblings with respect to staying on in education. As described above, Hong Kong was in general poor in that period. It was rather common for parents, especially those supporting a large number of children, to send their elder children to work at an early age so that the children would bring in financial contributions to the family (e.g. Hambro 1955). Mr. Wan[8] was one of those children, as he recalled:

(M)y father was quite pleased that I agreed to take up an apprenticeship in a factory after primary form six.… He had thirteen children to support. Even though he was a salesman, which was a relatively well-paid job at that time, you could not expect him to support every child to stay on at school.… Being the eldest child…making economic contributions to my family was my responsibility. (Mr. Wan, aged 55, with 12 siblings, of a middle-class origin: his father was a salesman)

4.3 What Mr. Wan meant by his responsibility of being the eldest for his family was not defined by any particular ideology but by a child’s financial obligation to the family commonly found in that period of Hong Kong (cf. Finch 1989), as Mrs. Yim articulated.

You can see there are two groups of siblings in my family – the eldest four are one group while the youngest five are the other. When the eldest four siblings were at school, my family was much poorer, whereas we, the youngest five, enjoyed the contribution from the eldest four, who started working in their teens. It was true that my elder siblings also got married early and didn’t make contributions to the family in a real sense after marriage. Yet they lightened my father’s burden and our immediate contribution was less urgently needed. So my father could let us, the youngest five children, study at school full-time. (Mrs. Yim, aged 35, with 8 siblings, of a working-class origin: her father was a factory worker)

4.4 What concerned EY parents was not upholding an ideology about proper roles of their elder children but, rather, upholding their immediate financial needs. With some elder children working and thus a lightened financial burden, EY parents could afford to support their younger children in education. Whether a younger child could be exempted from being sent out to work was contingent on the family finances at the time: whether the financial contributions of the existing working elder children were enough to secure the basic survival of the entire family. As in the case of Mr. Sung, the third of four children in his family, despite the financial contributions of his two elder sisters, he still had to quit schooling at the age of fifteen out of economic necessity; but for his contribution, his youngest brother might otherwise have had to work at an early age as well. In other words, EY parents were selective because they had such urgent economic needs, but not even because they were partial to their youngest children, as assumed in a Chinese folk belief (cf. Pieke 1991), or because they practised a particular kind of ideology about proper roles of the youngest. This is in line with the argument that the amount of resources available to the same parents changes with stage in the family-life-cycle (cf. Hogan and Kertzer 1988) and this shapes what they could do for the education of their elder and younger children. This lends further support to studies (e.g. Powell and Steelman 1993; cf. Freese at al. 1999) that refer to the number and the spacing of children in explaining different educational attainments of siblings. The experience of Mrs. Lei and her elder siblings experience echoes this.

My father’s income was not only unpredictable but also meagre…. My mother always said to me that as the youngest child, I was the most lucky in that I was allowed to stay on at each stage when I passed each public examination…. Some of my siblings were academically capable, but they were just unlucky in that my family was too poor to support them staying in education long enough to sit the examination. (Mrs. Lei, aged 35, with 7 siblings, of a self-employed origin: her father was a self-employed vendor)

4.5 One may wonder whether EY parents were selective because they aimed to have at least one child who could obtain a relatively high level of education. Having limited resources, it would have seemed sensible for parents to invest more of resources in the education of academically-more-capable children, because their chances of passing the necessary examinations would be higher. But, this was not what EY parents sought to achieve. Rather, they took their selective treatment as a strategy for basic survival. Their decision was not contingent on children’s academic ability but their financial situation at that point. When their elder children were at school, the finances of these parents were usually very tight; even if their elder children were academically very bright, they simply could not afford to let them stay on in education but had to treat them as an income source to make their ends meet. In contrast, when their younger children were at school, their financial burden became lighter; even if their younger children were not academically very bright, they could still afford to let them stay on as long as the younger children passed the required examinations. One more point is noted. Even if the financial situation had not improved at a later stage and parents wanted to send out their younger children to work, parents could perhaps still do it, albeit illegally, after 1964, when the act of legal working age set at fourteen was loosely enforced. But it was very difficult, if not impossible, for them to do it in the 1970s, when the government strongly enforced the act of compulsory basic education. In this sense, amongst the younger children, the youngest were most advantaged not only because more resources became available to the same parents at a later stage, but also because policies were changed to favour the children’s education. All these are in line with the finding reported in many studies that families acted collectively to advance the careers of not all but rather a selected child, such as the youngest child (cf. Elliott 1997).

SD parents

4.6 Whereas class differentials in educational attainment have persisted, a so-called ‘gender revolution’ has been achieved in many industrial-capitalist societies in that over the past twenty years girls have steadily outperformed boys at every level of education (Arnot et al. 1999). Freese and Powell (1999) even reported that nowadays girls in the United States, regardless of their social backgrounds, received more parental investment than did boys (cf. Behrman et al. 1995). While this ‘gender revolution’ has not been achieved even in nowadays Hong Kong, the traditional Chinese ideology described above seemed at work in the cases of SD parents. They considered that education was not for women and that it was prestigious for women to stay at home. In practising this ideology, what concerned the SD parents were their social status and their daughters’ prospects in the future marriage market. As an example, Mr. Lung’s father deliberately chose not to support the education of all his five daughters and kept them at home before marriage, because he saw doing so as a proper way of living up to the traditional Chinese ideology, as Mr. Lung proudly explained.

Of course, all my sisters stayed at home! You have to know, we are a family of intellectuals: all my ancestors are intellectuals and so am I. My father is a well-read traditional Chinese scholar. Why would he want his daughters to be seen in public before marriage? (Mr. Lung, aged 59, with six siblings, of a middle-class origin: his father was a school principal.)

4.7 In that period of Hong Kong, despite subscribing to this ideology, not every well-off parent could financially afford to practise it in this way. Then we would expect that keeping all their daughters at home before marriage was perhaps unthinkable for the poor. Yet, Mr. Doo’s father was a poor self-employed lighter-repairer but he still practised this ideology in the same way as the other SD parents. Given the nature of my data, the question of whether there is any correlation between class and practising the ideology in this particular way cannot be addressed here but should be explored further.

4.8 Expecting their daughters to advance through marriage and assuming that education was irrelevant to their marriage in the context of post-war Hong Kong, SD parents did not do much for their daughters’ education. In two cases, SD parents arranged marriages for their daughters. Despite having a disadvantaged educational career vis-à-vis her brothers, Mrs. Lung was quite happy about marrying into the middle class through her father’s arranged marriage.

How did you react to the marriage arranged by your father? Mr. Lung is a reliable and well-educated gentleman. And he had a decent teaching job. It is quite nice to have this kind of a husband. Moreover, his father was also a well-educated gentleman and was a school principal. So, why should I be unhappy with this marriage? (Mrs. Lung, aged 51, with 5 siblings, of an employer origin: her father was a merchant)

4.9 If resource availability explained why EY parents were selective, then ideology did the same for the cases of SD parents. Whereas EY parents were selective because they wanted to secure a basic survival for the entire family, SD parents were selective because they were strategic about the advancement of their sons, their daughters, and the whole family. As ideology played a strong role in structuring SD parents’ practices, their children’s academic ability was simply irrelevant to their discriminatory decisions.

SD-EY parents

4.10 As with EY parents, SD-EY were also concerned about basic survival, as Mr. Law explained.

My third sister, the eldest daughter, was the most unlucky among all of us for there were so many small children in the family. As far as studying at secondary school was concerned, the following description was true in the 1950s. Life was hard. When parents wanted to let their children study further, they were able only to let their sons do so after they had done their best. Daughters didn’t have this opportunity. (Mr. Law, aged 51, with 6 siblings, of a working-class origin: his father was a warehouse-keeper)

4.11 Perhaps economic concerns did play a part in the case of Mr. Law’s selective parents. However, one may wonder if they were selective purely out of economic necessity, why did they ask their eldest daughters but not eldest children, as EY parents did, to work at an early age? This suggests that the concern of SD-EY parents was more than securing a basic survival. While some of them were like SD parents seeing education as not for women, others considered that education was desirable for women but it was absolutely necessary for men; therefore, if a choice was to be made, the education of sons was always preferred, as indicated in the following quotations.

I got a place in a secondary school and asked my mother for the registration fee, which was only a few dollars. However, she refused to give me even a few dollars… She said my second brother was studying at secondary school, so she couldn’t let me do the same. Her reason was someone else would take care of me, whereas my second brother would have to take care of someone else. (Mrs. Tang, aged 42, with 5 siblings, of a working-class origin: her father was a factory worker)
My mother said that … daughters would get married some day anyway. If the daughters studied more and got higher qualifications, they would only benefit their future families. (Mrs. Yam, aged 50, with 11 siblings, of a middle-class origin: her father was a large proprietor)

4.12 Perhaps these could be seen as two rational and/or practical interpretations of the ideology. Some SD-EY parents found it sensible to invest more in a future breadwinner rather than a future dependant, and others found it rational to invest in the education of someone who would give them the most return in the future. If we consider that their desire of maintaining social status explained why SD parents practised the ideology in a so-called dogmatic way, then perhaps we could say that their financial concerns somehow led most SD-EY to practise the ideology in a practical and/or rational way. How SD-EY parents perceived their daughters’ prospects in the current and future labour markets might play a role in explaining their discriminatory decisions. In Hong Kong of the 1950s and 1960s, there was an abundance of low-paid manual jobs, especially unskilled ones in the manufacturing sector, which did not require any formal qualification; but there were only a limited number of well-paid advantaged clerical jobs, which required at least a secondary education. Subscribing to the ideology and thus assuming that their sons would support them in their old age, SD-EY parents might find it sensible to give their daughters a little education so that they could take up a manual job so as to make immediate financial contributions to them, and to keep their sons in education longer so that they would be able to take up a more well-paid advantaged clerical job and thus make greater financial contributions to them later on. In this sense, SD-EY parents’ decisions could be seen as strategic moves echoing Bourdieu’s (1990:148) metaphor of playing cards: in order to win the game (to maximise life-time financial gains from one’s children and bring social advancement to the entire family), the player should understand that the value of each move (the meaning of sending a daughter to work at an early age and of keeping a son in school) depends on the rules of the game (what is seen as a legitimate way of living up to the ideal of the traditional ideology) and the skill of playing (how well-informed they are about the current and future labour markets for men and women and how effective they are in making their sons and daughters do what they want them to do). It remained unclear whether SD parents differed from SD-EY parents because SD-EY parents were more liberal on gender issues, or because they could not afford to practise the ideology in the same way as did SD parents, or because they were more strategic than SD parents. But the observation of differences between SD and SD-EY parents per se highlighted the fact that the same traditional ideology could be practised in different ways in a so-called new or modern context (cf. Hareven 1978). Therefore, their children’s academic ability was irrelevant to SD parents’ discriminatory decisions but it could play a part in those of SD-EY parents. Yet, it worked rather differently for daughters and sons. SD-EY parents may not give their academically-bright daughters any support but they would definitely provide extra assistance for their academically-incapable sons, as Mrs. Luk recalled.

My family was quite well-off but…(w)e daughters have no status in the family. There would be no problems to let the sons complete a university education. But the daughters had to work right after secondary form five…. After passing the public examination, I begged my father to support my studies in a teacher training college…. My first sister…didn’t pass the public examination, so she even didn’t have a chance to beg…. My father forced my second brother to complete secondary form five even though he didn’t like studying. I think if he had been willing to study, my father would have sent him abroad…. My father wouldn’t have any expectations of the daughters. But he had straight demands that my two youngest brothers had to become doctors. (Mrs. Luk, aged 44, with 5 siblings, of an employer origin: her father ran a grocery store.)

4.13 In this sense, consistent with class effects reported in mobility studies, while being disadvantaged vis-à-vis their brothers, daughters of a middle-class origin (e.g. Mrs. Yam) or of an employer origin (e.g. Mrs. Luk) were still more advantaged than their working-class counterparts (e.g. Mrs. Tang) in that their parents were selective at a later stage. And therefore Mrs. Yam and Mrs. Luk could complete a secondary education – an advantaged qualification in that period of Hong Kong (Chan 1994) – and also a teaching diploma later on whereas Mrs. Tang could finish only a primary education. If resource availability shaped by class plays a role in deciding when SD-EY parents would become selective, then resource availability shaped by the family-life-cycle plays a part in deciding which sons and which daughters would be affected most by a selective parental treatment. Younger siblings are generally more advantaged than their elder siblings because at a later stage the finances of their parents are usually less tight. But at the end of the day, which daughter suffers most and which son benefits most from selective parental treatment is contingent on resource availability; this interactive effect of gender and birth order has not been fully captured by the variable of the number of siblings and that of spacing of children in quantitative studies, as the example of Mr. Cha illustrates.

Perhaps the people of the previous generation placed more importance on boys than girls. My parents considered that the girls didn’t need much education. I am the eldest son and all who followed me at that time were sisters. When I had finished a junior secondary education, my family was quite well-off and all my younger brothers were still at primary school. So it was very natural for my father to send me abroad. (Mr. Cha, aged 56, with 8 siblings, of an employer origin: his father was a merchant)

4.14 Like SD parents, SD-EY parents were selective because they were strategic about the advancement of their sons, their daughters, and the whole family. But, unlike SD parents, in addition to ideology, SD-EY parents were also responding to their children’s academic ability and resource availability, although each played a rather different role for sons and daughters. While the education of daughters was usually more expendable than that of their brothers, younger daughters or those academically-more-capable still had a better chance vis-à-vis their sisters to get some parental support, although there was no guarantee. While the education of sons was usually their parents’ top priority, elder sons or those academically-incapable-sons might lose out vis-à-vis their brothers when the finances of their parents were extremely tight. The effects of birth order and gender are not additional or separate but could be interactive. And these interactive effects seem even more clearly manifested in the cases of employer SD-EY fathers; a separate discussion of them is therefore required.

Employer selective parents

4.15 What the four employer SD-EY fathers had in common was that they had clear expectations of their children of different genders and birth orders. The following are two detailed illustrations:

4.16 The first case is Mr. Kwong’s father. Running a restaurant in England and having a farm in Hong Kong, he had four children: first son, second daughter, third son, and fourth son. He commanded his first son to quit schooling at the age of fifteen and to work in his restaurant in England, made his second daughter quit schooling at the age of twelve and work on his farm in Hong Kong, and encouraged his third and fourth sons to stay on in higher education. Getting his first son to work in his restaurant, Mr. Kwong’s father was preparing him for a future inheritance and also expected him to support the further studies of his youngest two sons; getting his second daughter to work on his farm, he simply wanted her immediate labour support for the family before marrying her off; getting his third and fourth sons to study further, he expected them to perform well so that they would eventually become professionals. In short, he expected his eldest son to become an employer taking over his business, his youngest two sons to become professionals, and his daughter to advance via marriage.

4.17 The second case is Mr. Yu’s father. Running an orthopaedic shop, he had eleven children. Mr. Yu’s father asked all his six daughters to stay at home after a primary education: they never had a paid job. But all his five sons completed at least a nine-year basic education. After the sons’ formal education, Mr. Yu’s father asked his fourth son to work in his orthopaedic shop, gave his second son start-up capital to set up a trading firm, and encouraged other sons to stay on in higher education. In brief, he expected one son to become an employer inheriting his trade, his academically-more-capable sons to pursue a professional career, his academically-less-capable son to take self-employment, and his daughters to advance via marriage.

4.18 Amongst all selective parents, employer parents were most specific about how they wanted their children to advance: marriage was for daughters, a professional career for academically-more-capable sons, self-employment for academically-less-capable sons, and inheriting the family business for one selected son, usually the eldest. This probably reflects two characteristics of the class position of employer parents: first, they have a family business to pass on to their children; and second, they could offer them a job. These two characteristics also lead mobility researchers to suspect that employer parents, as well as the self-employed, attach less importance to their children’s education (cf. Ganzeboom et al. 1991). However, the data here did not fully support this suspicion: employer parents at least wanted their academically-more-capable sons to attain a relatively high level of education and thus to become a professional. And, employer parents did not seem to take self-employment as a promising mobility strategy: they used it only for their academically-less-capable sons. This echoes Chiu’s (1998) finding about the self-employed in Hong Kong that they wanted their children to pursue a professional career and self-employment was only a second best option for them. And this is also consistent with the finding reported for many industrial-capitalist societies (e.g. Aldrich and Waldinger 1990; Aronson 1991) that self-employment is usually used by disadvantaged groups, especially immigrants, as a defence against unemployment.

4.19 However, what remains unclear is how these employer parents rank different mobility strategies for their sons. If they genuinely valued most a professional career, why did they not let all their sons pursue a professional career? Their decision could certainly be seen as a strategy of employer parents for preserving their family business. In order to avoid competition, and thus conflicts, among their sons, employer parents had to encourage all except one son to advance in other ways. Besides, consistent with one characteristic of Hong Kong that the size of business was particularly small (Sit and Wong 1989), the size of employer parents’ family business was actually too small to be divided among all their sons. Given the size of the family business, it also meant that inheriting their family business did not necessarily give a son a more promising future than self-employment. Since self-employment was taken as a second best mobility strategy, why, then, did employer parents still want to pass on their family business to one son? This decision seems to suggest that employer parents have reasons other than their sons’ advancement for desiring to preserve their own business over generations. This may also be related to how this inheritor son is selected. Is the selection based on ideology: parents choose the eldest son because they practise primogeniture, or they select the youngest son because they practise ultimogeniture? Or, is the selection based on merit: parents choose the son who is most talented in running the family business of that particular kind? Or, is the selection based on need, consistent with what economists call the principle of compensation instead of being neutral or egalitarian: parents choose the son who does not have other better mobility options and thus badly needs the inheritance? Having a family business to inherit could be a support and could also be a burden for a son, depending on the son’s ability and whether his career ambition and his father’s expectation of him coincide. It could be an attractive option for an academically-less-capable son compared with the prospects of becoming unemployed. But it could also be a burden for an academically-more-capable son preventing him from making a more promising career, as reported for those sons who had to quit their professional jobs in order to inherit their family farms in contemporary France (Bertaux-Wiame 1993). What this indicates is that compared with salaried parents, employer parents could play a much more direct role in their children’s education and career. This seems to suggest that, first, the educational and occupational attainments of children of employer parents could be even more diverse than those of children of salaried parents; and therefore, second, the role of employer parents in their children’s mobility should be taken more seriously so as to examine what kind of impacts they could have on the education and career of their different children.

Concluding remarks

5.1 An investigation of what selective parents were reported to have done for their children provided at least a partial answer to the puzzle posed at the outset: why do only some children attain a level of education considered to be socially desirable whereas their siblings do not? Perhaps it is true that they share the same class-specific culture; nevertheless children of the same family could develop rather different orientations to education. And, perhaps it is also true that they receive a similar amount of resources from their same parents; but some children still obtained more than their siblings with respect to their education. Differences in their children’s educational attainment could be seen as resulting mainly from deliberate decisions of the same selective parents. These were their strategies for basic survival or for advancement (cf. Moen and Wethington 1992). Consistent with findings reported in most mobility studies, resource availability, shaped by class as well as by the family-life-cycle, appears to play a significant role in the decisions of selective parents: it shapes when they would make such selective treatment and what they could do for their children’s education. But, resource availability alone does not tell us about what selective parents would do for their children’s education. The role of ideology should be seriously considered: although it is unclear how class and ideology are related or how an ideology would be practised, ideology somehow shapes what selective parents consider best for their different children and thus what they would do for the education of each child. As some sociologists put it, in choosing an educationally more ambitious option, working-class parents need a greater assurance than their middle-class counterparts from their children in terms of academic ability (e.g. Goldthorpe 1996). Similarly, in the socio-economic context described here, the same parents also need a greater assurance from some of their children than from the others. For daughters and/or elder children, being academically capable does not necessarily guarantee them any parental support. In contrast, for sons and/or younger children, being academically incapable could serve as a signal calling for more parental assistance. After taking account of resource availability, ideology, and children’s academic ability, we can say that the same selective parents are enhancing the education of their sons and/or younger children at the expense of that of their daughters and/or elder children. In short, the same parents could be viewed as playing a very constructive role in the education of their sons and/or younger children but at the same time could also be seen as disenabling that of their daughters and/or elder children. Contrary to our conventional wisdom, for some children their parents could well be the very first hurdle, if not the most difficult one, that they have to overcome in order to make educational advancement. Without taking variation in each family seriously, the positive and negative impacts of the same parents on their different children are likely to go unnoticed or would probably be offset by each other in aggregate statistical data. From my sample, which was constructed by selecting so-called successful people, we cannot be sure that every parent has explicit strategies for their children’s education, nor that such strategies, where they do exist, are always successful for the targeted children. Besides, there are factors other than the role of parents, positive or negative, in affecting one’s mobility trajectory and thus final mobility outcome. But the point here is this: parents in Hong Kong, especially employer parents, could play a direct and significant role in their children’s mobility and yet this has only been assumed but not seriously researched in mobility studies.

5.2 This examination also leads me to argue that an individual’s mobility should be examined within the context of the family and thus to support the stance that the family rather than an individual should be taken as the unit of analysis in mobility studies. The reason usually put forward by those who take this stance is that family members presumably share similar life chances. Unlike them, I take this stance not only because I take that presumption seriously but also because this stance gives rise to a more accurate conceptualisation of mobility. Mobility should not be conceptualised as purely an individual pursuit, since, as I have shown, the family is not simply one of many sources from which individuals can draw resources for making their social advancement but could be seen as an avenue of competition among siblings for parental resources and assistance. The mobility of the members of the same family is of an interdependent nature in that their mobility outcomes result from never-ending processes of negotiation among each other. An examination of inequality within the family could therefore provide important mechanisms explaining gender and class inequality in particular and social inequality in general. Having said that, I do not mean that the results reported here are definitive, even for the case of Hong Kong; rather, they should be treated as tentative, given the sample size and its nature. More empirical work is needed so as to examine to what extent the practices of these selective parents are specific to the Chinese in Hong Kong of that period and to what extent they could be expected to apply in other contexts of different periods.


1 This study was funded in the form of J.K. Swire Memorial Scholarship 1995-1998.

2 Targeted respondents should also have at least one child aged six or above. This selection criterion is irrelevant to this paper. For details for rationale behind selection criteria, see Wong (2001: chapter 2).

3 It is the major local dialect used in Hong Kong.

4 In terms of class origin defined by the occupation of respondents’ fathers, two were of a middle-class origin, two of a self-employed origin, and eight of a working-class origin.

5 One was of a middle-class origin, one of a self-employed origin, and one of a small-employer origin.

6 Four was of a middle-class origin, four of a small-employer origin, and seven of a working-class origin.

7 For the sake of convenience, I refer to the siblings of the respondents like second sister/daughter or third brother/son. Second sister/daughter means this sibling is the second child and she is a daughter, while third brother/son means this sibling is the third child and he is a son. But they may well be the eldest daughter or eldest son respectively in their families.

8 All names in this paper are fictitious.


Data for this article are from a study with which I earned my doctorate at Oxford. I am most grateful to John Goldthorpe and Gordon Marshall for supervising the study. As a COE fellow at the Center of the Study of Social Stratification and Inequality (CSSI), I benefit from its support of all kinds, including insights from Professor Yoshimichi Sato, the Director of CSSI, on the subject of social mobility. A big thank-you goes to CSSI and Professor Sato: without its support and his insights, this article would have been impossible. I also thank the Editors and two anonymous referees of Sociological Research Online for their very constructive comments and suggestions on earlier versions of this article. Finally, a special thank-you goes to Carter Johnson, who kindly made time to read the final draft of this article and gave me suggestions to improve it.


ALDRICH, H.E. and Waldinger, R. (1990) ‘Ethnicity and Entrepreneurship’ Annual Review of Sociology 16: 111-35.

ARNOT, M., David, M., and Weiner, G. (1999) Closing the Gender Gap: Post-War Education and Social Change, Cambridge: Polity.

ARONSON, R.L. (1991) Self-employment: A Labour Market Perspective, Ithaca, New York: ILR Press.

BAKER, H.D.R. (1979) Chinese Family and Kinship, London: Macmillan.

BEHRMAN, J.R., Pollak, R.A. and Taubman, P. (1995) From Parent to Child: Intrahousehold Allocations and Intergenerational Relations in the United States, Chicago: The University of Chicago Press.

BERTAUX-WIAME, I. (1993) ‘The Pull of Family Ties: Intergenerational Relationships and Life Paths’ in D. Bertaux and P. Thompson (editors) Between Generations: Family Models, Myths and Memories, Oxford: Oxford University Press.

BOUDON, R. (1974) Education, Opportunity and Social Inequality, New York: Wiley.

BOURDIEU, P. (1990) The Logic of Practice (translated by R. Nice), Stanford, Calif.: Stanford University Press.

BOURDIEU, P. and Passeron, J. (1977) Reproduction: In Education, Society and Culture (Translated by Richard Nice), London: Sage.

BURGUIERE, A., Klapisch-Zuber, C., Segalen, M. and Zonabend, F. (editors) (1996) (translated by S.H. Tension) A History of the Family, Cambridge: Polity Press.

CHAN, T.W. (1994) Social Mobility in Hong Kong, Oxford: D.Phil. Thesis, Nuffield College, University of Oxford.

CHIU, C. (1998) Small Business in Hong Kong: Accumulation and Accommodation, Hong Kong: The Chinese University Press.

ELDER, G.H. (1981) ‘History and the Life Course’ in D. Bertaux (ed) Biography and Society: The Life History Approach in the Social Sciences, Beverly Hill, Calif.: Sage Publications.

ELLIOTT, B. (1997) ‘Migration, Mobility, and Social Process: Scottish Migrants in Canada’ in D. Bertaux and P. Thompson (editors) Pathways to Social Class: A Qualitative Approach to Social Mobility, Oxford: Clarendon Press.

ENDACOTT, G.B. (1973) A History of Hong Kong, Hong Kong: Oxford University Press.

ERIKSON, R. and Goldthorpe, J.H. (1992) The Constant Flux: A Study of Class Mobility in Industrial Societies, Oxford: Oxford University Press.

FINCH, J. (1989) Family Obligations and Social Change, Cambridge: Polity Press.

FREESE, J. and Powell, B. (1999) ‘Sociobiology, Status, and Parental Investment in Sons and Daughters: Testing the Trivers-Willard Hypothesis’ American Journal of Sociology 106: 1704-43.

FREESE, J., Powell, B. and Steelman, L.C. (1999) ‘Rebel without a Cause or Effect: Birth Order and Social Attitudes’ American Sociological Review 64: 207-31.

GANZEBOOM, H.B.G., Luijkx, R. and Treiman, D.J. (1989) ‘Intergenerational Class Mobility in Comparative Perspective’ Research in Social Stratification and Mobility 8(1):3-84.

GOLDTHORPE, J.H. (1996) ‘Class Analysis and the Reorientation of Class Theory: The Case of Persisting Differentials in Educational Attainment’ British Journal of Sociology 47: 481-505.

GUO, G. and Van Wey, L.K. (1999) ‘Sibship Size and Intellectual Development: Is the relationship causal?’ American Sociological Review 64: 169-87.

HAMBRO, E. (1955) The Problems of Chinese Refugees in Hong Kong, Holland: Sijthoff.

HAREVEN, T.K. (1978) Transitions: The Family and the Life Course in Historical Perspective, New York: Academic Press.

HOGAN, D.P. and Kertzer, D.I. (1988) ‘Family Structure, Individual lives, and Societal Change’ in M.W. Riley (editor) Social Change and the Life Course, Newbury Park, Calif.: Sage Publications.

HONG KONG 1981: A Review of 1980, Hong Kong Government Publications Bureau, Hong Kong: Hong Kong Government Press.


- Hong Kong Statistics 1947-1967

- Hong Kong 1961 Population Census Main Report

- Hong Kong 1971 Population Census Main Report

HONG KONG REPORT FOR THE YEAR 1964, Hong Kong Government Press.

INGLEHART, R. and Baker, W.E. (2000) ‘Modernization, Cultural Change, and the Persistence of Traditional Values’ American Sociological Review 65: 19-51.

LEMAN, K. and Leman, K. (1987) The New Birth Order Book: Why You Are the Way You Are, New York: Dell Publishing.

MITCHELL, R.E. (1972) Family Life in Urban Hong Kong, Taipei: Oriental Cultural Service.

MOEN, P. and Wethington, E. (1992) ‘The Concept of Family Adaptive Strategies’ Annual Review of Sociology 18: 233-251.

MULLIGAN, C.B. (1997) Parental Priorities and Economic Inequality, Chicago: The University of Chicago Press.

PIEKE, F.N. (1991) ‘Chinese Educational Achievement and “Folk Theories of Success”’ Anthropology and Education Quarterly 22: 162-80.

POWELL, B. and Steelman, L.C. (1993) ‘The Educational Benefits of being Spaced Out: Sibship Density and Educational Progress’ American Sociological Review 58: 367-81.

SAVAGE, M. and Egerton, M. (1997) ‘Social Mobility, Individual Ability and the Inheritance of Class Inequality’ Sociology 31(4):645-672.

SIT, V.F.S. and Wong, S.L. (1989) Small and Medium Industries in an Export-Oriented Economy: The Case of Hong Kong, Hong Kong: Centre of Asian Studies, University of Hong Kong.

STEELMAN, L.C., Powell, B., Werum, R., and Carter, S. (2002) ‘Reconsidering the Effects of Sibling Configuration: Recent Advances and Challenges’ Annual Review of Sociology 28(1):243-269.

SULLOWAY, F.J. (1996) Born to Rebel: Birth Order, Family Dynamics, and Creative Lives, New York: Pantheon Books.

VAN EIJCK, K. and De Graaf, P.M. (1995) ‘The Effects of Family Structure on the Educational Attainment of Siblings in Hungary’ European Sociological Review 11: 273-92.

WALBY, S. (1990) Theorising Patriarchy, Oxford: Blackwell.

WONG, Y.L. (2001) Family Strategy: A Study of Intergenerational Mobility in Hong Kong, Oxford: D.Phil. Thesis, Nuffield College, University of Oxford.

ZAJONC, R.B. and Markus, G.B. (1975) ‘Birth Order and Intellectual Development’ Psychological Review 82: 74-88.