The dangers of drawing cohort profiles from period data: a research note

Alyson A. van Raalte; Ugofilippo Basellini; Carlo Giovanni  Camarda; Marília R. Nepomuceno; Mikko Myrskylä

doi:10.31235/osf.io/frkcw

The dangers of drawing cohort profiles from period data: a research note

Alyson A. van Raalte, Ugofilippo Basellini, Carlo Giovanni Camarda, Marília R. Nepomuceno, Mikko Myrskylä

Centre for Social Data Science, CSDS

Research output: Working paper › Preprint

Abstract

Drawing cohort profiles and cohort forecasts from grids of age-period data is common practice in demography. In this research note, we: (1) estimate the bias in the cohort TFR and life expectancies calculated from such data, and (2) demonstrate that cohort Lee-Carter forecasts drawn from an age-period grid have implausible prediction intervals, especially when drawn from the upper and lower bounds of the time series parameter. These biases are surprisingly large, even when the cohort profiles are created from single-age, single-year period data. The danger is that we overinterpret deviations from expected trends, that in actual fact were induced by our own data manipulation.

Original language	English
Place of Publication	Rostock
Publisher	SocArXiv
Number of pages	21
DOIs	https://doi.org/10.31235/osf.io/frkcw
Publication status	Published - 3 Mar 2022
MoE publication type	Not Eligible

Fields of Science

5200 Other social sciences

Access to Document

10.31235/osf.io/frkcw

Cite this

@techreport{2fa3f60780974e5badc2b5f7de271e2f,

title = "The dangers of drawing cohort profiles from period data: a research note",

abstract = "Drawing cohort profiles and cohort forecasts from grids of age-period data is common practice in demography. In this research note, we: (1) estimate the bias in the cohort TFR and life expectancies calculated from such data, and (2) demonstrate that cohort Lee-Carter forecasts drawn from an age-period grid have implausible prediction intervals, especially when drawn from the upper and lower bounds of the time series parameter. These biases are surprisingly large, even when the cohort profiles are created from single-age, single-year period data. The danger is that we overinterpret deviations from expected trends, that in actual fact were induced by our own data manipulation.",

keywords = "5200 Other social sciences",

author = "{van Raalte}, {Alyson A.} and Ugofilippo Basellini and Camarda, {Carlo Giovanni} and Nepomuceno, {Mar{\'i}lia R.} and Mikko Myrskyl{\"a}",

year = "2022",

month = mar,

day = "3",

doi = "10.31235/osf.io/frkcw",

language = "English",

publisher = "SocArXiv ",

address = "United States",

type = "WorkingPaper",

institution = "SocArXiv ",

}

TY - UNPB

T1 - The dangers of drawing cohort profiles from period data

T2 - a research note

AU - van Raalte, Alyson A.

AU - Basellini, Ugofilippo

AU - Camarda, Carlo Giovanni

AU - Nepomuceno, Marília R.

AU - Myrskylä, Mikko

PY - 2022/3/3

Y1 - 2022/3/3

N2 - Drawing cohort profiles and cohort forecasts from grids of age-period data is common practice in demography. In this research note, we: (1) estimate the bias in the cohort TFR and life expectancies calculated from such data, and (2) demonstrate that cohort Lee-Carter forecasts drawn from an age-period grid have implausible prediction intervals, especially when drawn from the upper and lower bounds of the time series parameter. These biases are surprisingly large, even when the cohort profiles are created from single-age, single-year period data. The danger is that we overinterpret deviations from expected trends, that in actual fact were induced by our own data manipulation.

AB - Drawing cohort profiles and cohort forecasts from grids of age-period data is common practice in demography. In this research note, we: (1) estimate the bias in the cohort TFR and life expectancies calculated from such data, and (2) demonstrate that cohort Lee-Carter forecasts drawn from an age-period grid have implausible prediction intervals, especially when drawn from the upper and lower bounds of the time series parameter. These biases are surprisingly large, even when the cohort profiles are created from single-age, single-year period data. The danger is that we overinterpret deviations from expected trends, that in actual fact were induced by our own data manipulation.

KW - 5200 Other social sciences

U2 - 10.31235/osf.io/frkcw

DO - 10.31235/osf.io/frkcw

M3 - Preprint

BT - The dangers of drawing cohort profiles from period data

PB - SocArXiv

CY - Rostock

ER -