« Banjos, cricket and Wodehouse | Main | Cognitive biases against freedom »

February 15, 2006



Can you exclude the move to 'short-term sick' or 'long-term sick' that easily. THe major change in early retired is for women, down 11% on the quarter, and both categories of 'sick' have risen for them.

angry economist

I'd say that its inaccurate to claim this from your analysis - its a bit of a stab in the dark to link LFS figures to stock market wealth.

It would be best to analyse changes early retirees as a proportion of the workforce.

Stating the increase or reduction in stock can only be made when appreciating the size of the labour pool.

i.e. x% of working age men are in early retirement - is this higher or lower than Q2 etc.

Plus- these figures are obviously from the Labour Force Survey - this kind of stock change (39,000) could partly be due to sampling error. e.g. the usual error of margin for LFS (non-income data) is 10,000 either way.

Plus - seasonal effects in Q4 - some casual labour might go into work at xmas and come out of retirement - e.g. think of all those santa clauses in our shopping malls at that time of year.

As a rule, to be thorough with the LFS, I would pick a quarter (best usually Q2) and analyse every year over past 5-10 years.

I wish I had time to look at the LFS returns for you but alas I am busy working all hours doing a much more meaningless task.


If you look at the activity rate of the population over retirement age in the LFS the response is even more impressive. The number of those over pensionable age active in the labour market is growing at around 10% a year. The activity rate of that group has risen from 9.4% in Q4 2003 to 10.5% in Q4 2005 (the activity rate for the rest of the population was largely unchanged at 78.6% over the same period).

angry economist

well done Neville! now what could explain it?

there might be push factors, such as pensions crisis.

but there might be pull factors - such as about 9 years of bouyant labour demand and impending (or is it now law?)legislation outlawing age discrimination in the workplace and in recruitment

I am not sure whether there are questions about pensions etc in the LFS returns. Don't recall this.


Given that employment growth outside the 60/65+ bracket was actually quite low in the past year and associated with subdued nominal and real wage growth, and the number of unemployed is up on the year, I think it's quite hard to argue that the rise in employment (and activity) of the 60/65+ cohort is demand driven.


Angry - your point about sampling error is good. Your point about seasonality is less so, as the data are seasonally adjusted, and there seems little residual seasonality; this winter's fall seems abnormal.
Sure, it's not just the stock market that affects retirement. Another consideration is annuity rates. Maybe the latest drop in early retirement is partly due to the fall in these.
Again, though, what's impressive in the longer-term data is that early retirenet stayed flat in the face of the drop in these rates since the late 90s.
Neville - I looked at early retirees rather than over-65s because their work-retirement choices should be more sensitive to the drop in the stock market as:
1. People in work in their 50s have the choice to retire or not, so postponing retirement would show up in these data, not data for over 65s.
2. Those who have taken early retirement might be more attractive to employers, and have better health, than older workers, and so would be more able to get work if their retirement income was inadequate.
3. They should be more exposed to the stock market fall of 2000-03. Older people would have had more earlier gains as a cushion and/or should have been more likely to switch to bonds as they age (though there's all sorts of issues raised here...)

angry economist

Yeh Chris you are right about seasonality being smoothed out but I have also just remembered that all quarterly figures are 'annualised' - they take rolling averages of the 4 quarters. This makes my seasonality point redundant but might make your points more salient rather than less so- i.e. we are not looking at a Q4 spike but an annualised change.

If I had more time I could help answer questions 1. to 3. as I have access to LFS returns on my desktop and can go through them with SPSS. Fairly easy to profile those taking early retirement in the UK, including their previous role, remuneration etc, as well as profile those still in work of a certain age, or working beyond 65. I will put it on the shelf as something to do when I have a little down time at work.


There are a couple of local authority deals which offered much less generous terms for early retirement (and a cohort of teachers) which ought to be coming into the statistics about now.

The comments to this entry are closed.

blogs I like

Blog powered by Typepad