본 연구는 신규 시행되는 기초연금이 노인의 소득 보장에 기여하는 바를 시뮬레이션으로 검증하였다. 한국복지패널의 자료를 활용하여 기초연금 수급의 상황을 가상으로 구성한 후, 이중차이분석 및 삼중차이분석을 사용하여 기초연금 비수급집단과의 소득 차이를 분석하였다. 이때 성향점수 매칭 방법을 결합함으로서 비교의 대상이 되는 두 집단 간의 유사성을 확보하였다. 첫째, 노인의 성별·연령·가구유형에서는 각 특성의 단면이 적용된 이중차이분석만이 유의하였으며 삼중차이분석은 기각되었다. 둘째, 노인의 소득 및 지출 관련 요인을 고려한 분석에서도 월 평균 지출이 90만원 이하인 집단을 제외한 모든 집단내에서 기초연금의 효과가 유의하였으며, 삼중차이분석(DDD)의 유의한 결과는 경상소득·지출 변인에서만 추정되었다. 한편 국민연금 및 국민기초생활보장급여 각각 모두 비수급집단에서만 기초연금의 효과가 유의하며 수급집단에서는 그렇지 않은 것으로 확인되었다. 또한 양자를 투입한 삼중차이분석(DDD) 추정치 역시 유의수준의 범위를 벗어나, 기초연금과 국민연금 및 기초생활보장제도 간의 관계성이 통계적으로 유의하지 않은 것으로 나타났다.

This study examines how effective the basic pension scheme, implemented in July 2014, has been by applying a simulation method. First, we identified the status of entitlement given to individual senior citizens and calculated the monthly basic pension income of the identified pensioners based on the data from the Korea Welfare Panel Study (KoWePS). This was followed by a comparison of income changes between the beneficiaries and non-beneficiaries with an aim to analyze how the scheme contributed to economic lives of senior citizens. The methods of analysis include Difference-in-Difference (DID) and Difference-in-Difference-in-Differences (DDD). Particularly, the latter was applied to identify the interaction mechanism between the characteristics of older people and the basic pension system. They were combined with the Propensity Score Matching (PSM) to reduce the selection bias and to secure the similarity between both groups. First, DID results consider gender, age and household are significant while their DDD results are not discarded. Second, in terms of the DID results based on the analysis of the factors related to the income and expenses of older people, the basic pension is found effective for all groups but for the group whose monthly average expense is lower than 900,000 KRW. However, the significant result from DDD was only estimated from the ordinary income expense factor. Third, both national pension and national basic living security payment are effective for the non beneficiary group of the basic pension while they are not with that of the beneficiary group. It supports those discussions which has been continuing regarding their combinations from the institutional level and we are expecting further examinations in the future. Also, their estimates applying DDD approach are not significant, resulting in a conclusion that correlations of the three schemes were not statistically significant. Based on the results, all three factors, individual and household characteristics of older people are not significantly related to the basic pension scheme. Meanwhile, income and purchasing power of individual older people are the only factors that are systemically related to the basic pension. Moreover, the final result supports the needs to review and improve effectiveness of the basic pension scheme and the whole old-age income security systems, raising a concern of mass production of welfare contradictions where welfare benefits are concentrated to a specific group. The result and implication of this study will be able to be discussed in more detail based on the empirical and logical reasons after the future system implementation. Furthermore, it is required to understand the characteristics and the following needs of the older people from the institutional level, in order to improve effectiveness and sustainability of the basic pension system and further improve the soundness of the entire social security system.