ソーシャル・キャピタルを測る <論文>

Social capital in the United States of America was quantified in the present study using official statistics in latent factor analysis. Social Capital is defined in this investigation as the degree of stable governance for the following: ① safety and security, ② wealth in both spiritual and material terms, and ③ social anticipation and networks. Social capital was thus measured as a single common latent factor that affected the degree of performance through such data as ①the number of murders, and suicides and unemployment level, ② personal income and state expenditures on education and welfare, and ③ the number of Nonprofit organizations (501 (c)(3)] and charity donations.　　 From the above analysis, social capital was confirmed as a single common latent factor that exerted a direct effect on social statistics for the years, 1995 (Fig. 1 (b)), 2005 (Fig. 1 (a)) and 2010 (Fig. 1 (c)). The multiple regression coefficient for the number of nonprofit organization was nearly 1.0 in most cases. This means that variation in the number of nonprofit organizations was almost completely determined by population size and social capital. Thus, the regression-based scores (regression residues) for the number of nonprofit organizations can be used as a proxy statistic for social capital.　　 The direct negative effect between the number of nonprofit organizations and murders was detected with statistical significance at the 5% level for the years, 1995 (Fig. 2(b)), 2005 (Fig. 2(a)) and 2010 (Fig. 2(c)). In addition, a direct negative effect was evident for both robberies (Fig. 4(a), Fig. 5(a)) and burglaries (Fig. 4(b), Fig. 5(b)). Using Putnam's data on "Trust" and "Honest", I confirmed a significant direct negative effect on murder (Fig. 8).　　 The results of the above analysis are consistent with those of Putnam (Putnam 2000) (Figs. 6, 7).