Thank you! Very interesting!
1. “use a Sacher estimate”
The curvature of the qx I obtain is greater (/starts at earlier ages) than what would be expected with a Gompertz Law (your graph on p16 of http://ift.tt/1yjXyCW ) so there might be something. I will use a Sacher estimate and will spread the message around me.
2. interesting points IMHO for GRG readers:
— GRG data
You refer to pionneering papers that found some mortality plateau for supercentenarians, before the technical issues that you highlight were so highlighted. You do not mention the analysis of GRG data with a Sacher estimate; or have I
— clarity on HMD smoothing
I understand that instead of finding the information at an official place, someone told you that “these and these but not those” cohorts were raw data or smoothed data (p4 you write “Data for
1900 and later birth cohorts were smoothed using Kannisto formula (Jdanov, personal communication)”) ==> given the high number of people who use HMD data, do you think it would be difficult to have somewhere online the list of HMD data that is smoothed or not? whom to ask?
Very best regards
PS: just for the record, I re-attach the HMD curves (Swiss
data being normal) and accordingly rectify my email below
De : Leonid Gavrilov À : Gerontology Research Group Cc : Natalia Gavrilova ; Edouard Debonneuil Envoyé le : Dimanche 10 août 2014 5h28Objet : Re: [GRG] New Study of Longevity Risk
Thank you for your interesting reply. The Society of Actuaries will
publish a peer-reviewed monograph soon, which will address your
2014 Living to 100 Monograph
http://ift.tt/1nEivCr, look for our book chapter there, when it is
published:Mortality Trajectories at Extreme Old Ages: A Comparison Study of
Different Data Sources on Old-Age MortalityIf you can not wait for final official publication, a preliminary
draft of this chapter is available here:
http://ift.tt/1spMdmq, you may find interesting Appendix A at pages 15-18
Hope it helps.
We will have a long international scientific travel starting this Monday,
and we will be out of office with limited e-mail connection until
So we can discuss the remaining questions when we return back in