Social capital works on whole networks and its process (chance of investment and return) is structured along with the configuration of them. However, in an ordinary survey design, it is quite difficult to locate respondents' personal networks on such macro configurations. Here we introduce a concept of whole-net base. It is as a shared attribute that indicates a latent whole network behind it. Keeping contact with a schoolmate, for example, indicates that the tie is a part of a whole network that consists of latently all graduates of the school. Under a question how social capital is built in the stratification process, we analyze a personal network data from the viewpoint of investment-return through various friends-net bases. The results show that; 1) the upper stratum invests and gets return more than the lower, especially on job-net and school-net bases that directly correspond to socio-economic status, however, 2) it is the lower stratum in which efficient investment (the tendency that more investment correspond to more return) is observed clearly, moreover, the observation is not limited to the specific whole-net base. These suggest a dynamic opportunity structure of social capital: while it reinforces the advantageous socio-economic positioning for the upper stratum, at the same time, it also makes their life chances greater for the lower stratum. However, we have to pay attention to those elder people in the lower of lower strata who do not have any access to social capital.

抄録全体を表示

In this paper, we focused on trust as one of the factors of social capital, and examined its generating processes by comparing two main theoretical approaches of trust — the Emancipation Theory of Trust and the Reduction Approach. The former theory states that creation of general trust will be undermined when commitment relations with specific others are strong, whereas the latter approach states that strength of commitment relations with specific others provides the basis for nurturing general trust. First we examined the relationship between city size (supposedly an indicator of the strength of committed relations) and the level of general trust using data from two social surveys conducted in Japan. The result of the analysis showed no correlation between city size and the level of general trust. Then we performed structural equation modeling to explore the relationships between a wider variety of capitals, social resources and general trust. The results of the analyses show that general trust is nurtured when it is based on relations of trust with specific people. In other words, the result is diametrically opposed to the Emancipation Theory of Trust in that it implies that placing a high priority on existing relationships with people fosters accumulation of general trust and therefore of social capital.

抄録全体を表示

It is empirically verified that people with higher income are, on an average, more trustful than those with lower income. If this finding is unconditionally valid, a dismal society may emerge and lead to the rich becoming richer by trusting strangers and receiving higher payoffs. The purpose of this paper is to explore the conditions that facilitate or hinder the growth of a dismal society by developing an agent-based model. We examined the hypothesis that the attractiveness of the market facilitates the trusting behavior of the rich, which results in a wider gap of the accumulated payoffs between the rich and the poor. In the model, agents decide whether or not they should leave their assurance group in order to receive a higher payoff from a stranger in the market. If an agent's partner is trustworthy, the agent receives a higher payoff from the partner than from a member in his/her assurance group. However, if the partner is not trustworthy, the agent receives a lower payoff. After receiving payoffs, agents learn to modify their behavior. The result of the simulation reveals that the gap between the rich and the poor widens only in the case of a moderately attractive market.

抄録全体を表示

In this paper, we discuss one of the major factors in social capital, the issue of trust, in terms of detection of trustworthiness of others. Yamagishi's Emancipation Theory of Trust concluded that high trusters, or people who have a high level of general trustfulness toward others, have social intelligence to accurately detect the general character of others. Kikuchi, Watanabe, and Yamagishi's Detection Experiment provided the empirical basis of such a conclusion. However, the Prisoner's Dilemma (PD) Game adopted in the Detection Experiment did not necessarily use appropriate alternatives of behaviors for participants as indicators of the general characters of people. Consequently, the results from the experiment are not appropriate for reviewing the ability to “detect a particular person's general character.” In order to overcome such problems and to measure the ability to detect a person's trustworthiness more appropriately, we conducted a laboratory experiment by adopting the game of enthronement. After a series of analyses, we could not confirm such a relationship between trustfulness and the ability to detect trustworthiness as discussed in a series of studies by Yamagishi. On the other hand, after reviewing the relationship between breadth of beliefs about others and the ability to detect trustworthiness of others based on Kelley and Stahelski's Triangle Hypothesis in a traditional PD game study, we could confirm that it is a person's assumptions or beliefs about internalization of social norms by others that determine the accuracy of his/her ability to detect trustworthiness of others.

抄録全体を表示

For the last two decades, there has been a growing argument that “social capital” derived from human relationships has an influence on various aspects of social phenomena. In many such studies, it is pointed out that social capital has two aspects social relationships as “network” and “trust” associated with such relationships. However, there has been a tendency for them to be discussed separately and they have not been integrated into one concept of social capital.
In this paper, we would like to define the ‘tie’ of human relationships and ‘strength of tie’ among the component concepts of social capital in a formal manner, and try to define the ‘trust’ mechanism as a function that maps ‘one's act counting on other people's actions’ into its feasibility. We will then explore the empirical credibility of such formalization by analyzing survey data.
In the following two sections, we will make clear related problems. First, we try to distinguish the ‘feasibility of one's acts which count on other people's actions’ from the probability of other people's action. In our view, one will never calculate an expected utility using the probability of other people's actions in everyday life. He/she will only judge whether his/her act counting on other people's actions is feasible or not. In the next section, we will distinguish the ‘trust’ mechanism from the ‘confidence’ mechanism according to Luhmann. We see their difference as a difference in domain of feasibility that each mechanism concerns.

抄録全体を表示

数理社会学は, 社会学の一分野であるという意味で, 経験科学に属する. 本稿ではこの基本的な認識を確認しつつ, FKモデルにおけるモデルの検証にかんする問題を指摘する. 具体的には, モデルの初期条件となる「客観階層システム」の想定の仕方によって, モデルと経験的データとの適合度が変わるという問題を指摘し, 理論的に検討を加える. モデルの検証についての厳密な検討を経ることによって, FKモデルの, ひいては数理社会学のより一層の発展が期待される.

抄録全体を表示

浜田・石田 (2003) は, J. ローマーの「機会平等の原則」というアイデアに基づき, 性別や親の地位などの機会の差を仮想的に調整した社会のジニ係数を分析する方法を定式化した. しかしながら, 浜田・石田 (2003) では, ローマー・モデルと仮想的機会調整分析法の仮定の違いが明確に規定されておらず, 機会不平等調整前後のジニ係数の差が統計的に有意であるかどうかも考慮されていなかった.
そこで本稿では, ローマーの規範的モデルと仮想的機会調整分析法の違いを明確化しつつ, ブートストラップ法を応用することにより, 機会調整前後のジニ係数の有意差検定を行う手法を提唱する. さらに, 機会変数が比率尺度である場合の分割数に関する問題についても言及し, 分析法の理論的性質を明らかにする. また分析例としてSSMデータを用い, 不動産相続額を機会変数とみなして, 所有不動産額と世帯所得という結果の配分への影響を検証する.

抄録全体を表示