This thesis studies the mixed data types kernel estimation framework for the models of discrete dependent variables, which are known as kernel discrete conditional functions. The conventional parametric multinomial logit MNL model is compared with the mixed data types kernel conditional density estimator in Chapter (2). A new kernel estimator for discrete time single state hazard models is developed in Chapter (3), and named as the discrete time “external kernel hazard” estimator. The discrete time (mixed) proportional hazard estimators are then compared with the discrete time external kernel hazard estimator empirically in Chapter (4). The work in Chapter (2) attempts to estimate a labour force participation decision model using a cross-section data from the UK labour force survey in 2007. The work in Chapter (4) estimates a hazard rate for job-vacancies in weeks, using data from Lancashire Careers Service (LCS) between the period from March 1988 to June 1992. The evidences from the vast literature regarding female labour force participation and the job-market random matching theory are used to examine the empirical results of the estimators. The parametric estimator are tighten by the restrictive assumption regarding the link function of the discrete dependent variable and the dummy variables of the discrete covariates. Adding interaction terms improves the performance of the parametric models but encounters other risks like generating multicollinearity problem, increasing the singularity of the data matrix and complicates the computation of the ML function. On the other hand, the mixed data types kernel estimation framework shows an outstanding performance compared with the conventional parametric estimation methods. The kernel functions that are used for the discrete variables, including the dependent variable, in the mixed data types estimation framework, have substantially improved the performance of the kernel estimators. The kernel framework uses very few assumptions about the functional form of the variables in the model, and relay on the right choice of the kernel functions in the estimator. The outcomes of the kernel conditional density shows that female education level and fertility have high impact on females propensity to work and be in the labour force. The kernel conditional density estimator captures more heterogeneity among the females in the sample than the MNL model due to the restrictive parametric assumptions in the later. The (mixed) proportional hazard framework, on the other hand, missed to capture the effect of the job-market tightness in the job-vacancies hazard rate and produce inconsistent results when the assumptions regarding the distribution of the unobserved heterogeneity are changed. The external kernel hazard estimator overcomes those problems and produce results that consistent with the job market random matching theory. The results in this thesis are useful for nonparametric estimation research in econometrics and in labour economics research.

本篇論文應用Rubin (1973)、Rosenbaum and Rubin (1983、1985a,b)所發展的配對方法(Matching Method)探討兩個議題，其一是評估企業從事社會責任行為是否可以改善績效，另一則是探討銀行加入金控後對其績效的影響。 / 針對第一個議題，我們將FTSE All-share 指數中被收錄於英國FTSE 社會責任指數的成份股定義為社會責任企業，未被收錄者定義為非社會責任企業，應用配對方法，根據特性變數是否相近為依據進行樣本配對。實證結果發現，不論使用配對前或是配對後的樣本進行分析，皆找不到社會責任企業的平均績效優於非社會責任企業的證據；相反地，部分的證據顯示社會責任企業的財務績效相對較低，因此我們的分析結果傾向支持焦點移轉假說，社會責任行為對財務績效的影響效果為負。 / 另外，我們以遠見雜誌於2006 年公佈的企業社會責任調查為基礎，將衡量企業社會責任績效的三個層面－社區參與、環境保護與財務透明，透過投資組合分析與迴歸分析來評估公司承擔社會責任與財務績效之間的關聯性。實證結果發現，第一，平均來說，公司的社會責任評分愈高，在評比結果公佈後的股票報酬愈低，兩者之間呈現負向關係；第二，公司在不同層面的社會責任行為對於股價報酬的影響具有差異性；第三、透過公司的長期財務資料發現，社會責任評分高的公司其長期平均會計績效相對較佳，但長期平均的股票報酬相對較低，表示社會責任型公司不一定是一個好的投資標的；最後，我們找不到支持社會責任行為可以做為公司績效保險的證據。本文使用各別層面的評比指標，更廣泛地評估社會責任行為對財務績效的影響，改進既有文獻中僅使用單一指標的不足之處。 / 就第二個議題而言，同樣地應用配對方法，根據銀行的特性變數是否相近為依據進行樣本篩選，利用台灣上市上櫃的銀行在2002 年第1 季至2006 年第2季的資料，我們發現在樣本配對前，平均來說，金控銀行在15 個CAMEL 績效指標的表現上皆相對較佳；樣本配對後，金控銀行除了在費用比率之外，其他的評比指標仍相對優於獨立銀行；不論使用配對前或配對後的樣本做分析，皆未出現金控銀行績效顯著較差的證據，因此，我們的實證結果偏向支持綜效假設，銀行加入金控對績效的影響效果為正向。 / 我們的研究架構與方案評估與政策衝擊分析文獻中的方法相一致，而貢獻則在於我們考慮樣本之間特性變數的差異性，進而以樣本配對的方式修正這個差異性，將有效降低既有文獻中在檢驗上述兩個議題上的選擇偏誤問題。 / In this thesis, we apply several matching methods, develop by Rubin (1973)、Rosenbaum and Rubin (1983, 1985a,b), on examing the effects of Corporate Social Responsibility (CSR) on firm financial performance (CSR-effect) and on evaluating the effects of banks being subordinated to FHC on CAMEL indicators (join-FHC effect). / For the former application, two competing hypotheses, social impact hypothesis and shift of focus hypothesis, are proposed, where the former suggests that CSR has a positive relation with performance and the latter are opposite. To ensure the CSR-effect is not contaminated by other factors or the sample is randomly drawn, we employ four matching methods, Nearest, Caliper, Mahala and Mahala Caliper to match the characteristics of the firms with CSR (CSR-firms) and without CSR (NonCSR-firms) to get rid of sample selection bias. Although the four methods yield slightly different results, to our surprised, firms engaging in CSR activities obtain lower values on the return on assets, return on equity, return on sales and earnings per share. Therefore, CSR at the very least does not improve the firm performance, supporting the shift of focus hypothesis. Engaging in CSR activities lead to more pain than gain, at least in the short run. / On the CSR topic, using a set of disaggregated social performance indicators for community participation, environmental protection and fiancail transparency from the Global Views Monthly, we examine the relationship between CSR and financial performance of TSE listing companies. Our main results show that first, scores on composite social performance indicators are negatively related to stock returns and this relationship cannot be rationalized by multi-factor models for explaining the cross-sectional variation in stock returns. Second, the effects of three disaggregated social performance indicators on stock returns are differenct among industries. Third, aggregated social performance indicator is positively related to long-term accounting performance but negatively related to long-term market performance, and thus implies that good companies are good in books, but not good investments. Finally, firms with high CSR ratings do no exhibit smaller decline of their financial performance during stagnations and thus not support the view that CSR is insurance of financial performance. / The second application, two hypotheses, synergy hypothesis and specialization advantage hypothesis, are raised too, where the former suggests that FHC-banks performer better and the latter are contrary. Above four matching methods are still used to fix the characteristics of two groups of banks in order to correct for sample selection bias. Based on after-matching samples, most of our empirical results suggest that FHC-banks, on average, outperform independent banks on most of indicators on capital adequacy, asset quality, earning ability and liquidity but not on management ability and being FHC-banks at the very least does not deteriorate the performance of banks, making our conclusion favors the synergy hypothesis and against specialization advantage hypothesis in Taiwan. / Our framework is identical to many applications of matching method such as Persson (2001), Hutchison (2004), Glick, Guo and Hutchison (2006), and are also parallel to standard medical and biological research.