Evaluating the Performance of Variants of the Lee-Carter Method for Mortality Forecasting in the Population Projection of the Czech Republic from the Czech Statistical Office 2023–2100

David Morávek

Evaluating the Performance of Variants of the Lee-Carter Method for Mortality Forecasting in the Population Projection of the Czech Republic from the Czech Statistical Office 2023–2100

Číslo: 4/2024
Periodikum: Demografie
DOI: 10.54694/dem.0353

Klíčová slova: mortality forecasting, Lee-Carter method, European mortality potential, Czech Republic

Pro získání musíte mít účet v Citace PRO.

Přečíst po přihlášení

Anotace: This article discusses the use of the Lee-Carter model, one of the most widely used methods for mortality forecasting. It also discusses its modified versions, namely the cohort and coherent variants, which were developed to improve the forecast accuracy of the model. This article compares them in terms of their ability to predict long-term mortality trends in the case of the Czech Republic. The projection of mortality was based on historical data on age-specific mortality rates from 1980 to 2019 for the so-called European mortality potential, which was constructed from the lowest observed mortality rates in developed European countries. On the basis of the results, the coherent Lee-Carter model was selected as the most appropriate model, which is described by the literature as more robust and suitable for projections with a long-term horizon.