Demographic Components of Future Population Change and Their Implications

Abstract

Low fertility and future population change Korea’s fertility rate started going down below the replacement level in the early 1980s and kept falling to a greater extent in the 2000s and afterward. With a view to addressing the ramifications that might arise from low fertility, the Korean government embarked in 2006 on the Basic Plan on Low Fertility and Aging Society, which is in its third 5-year iteration since 20161. Korea’s protracted low fertility has raised intense social concerns, but the long-term implications it has for future population changes have yet to gain enough attention. Indeed, the Third Basic Plan on Low Fertility and Aging Society has among its goals raising the total fertility rate to the replacement level (2.1) by 2045. What we need at this point is to consider whether the demographic problems Korea is facing can be resolved once and for all by raising fertility rate, if not whether the fertility target as stated in the Third Basic Plan is attainable at all. This study is aimed at examining how more than three decades of sub-replacement fertility will affect population change, in terms in particular of depopulation and population aging. Looking at the growth and decline of the population as forecast in Statistics Korea’s Population Projections (medium variant projection of 2016), this study employs a decomposition method that quantifies the contribution of fertility, mortality, migration, and age-structure factors to population changes. Additional simulations are used to examine how population changes will play out in the long run.2 Statistics Korea’s Population Projections (2016), the first of its kind to cover a period as expansive as 100 years (2015~2155), provides a useful reference point showing a long-term process of population change.3 In my reference scenario, I used the values as estimated in the 2016 Population Projections. In estimating the contribution of key factors associated with population change, I used the “cohort-component method.”