Jan 242013

These are comments on “Spending Power Across the Generations” by David Kingman, published on 23 January 2013 by the Intergenerational Foundation. I was directed to the report as follows:

Shiv Malik tweet

In summary:

  • It is about “spending behaviour”, not “spending power”.
  • It it restricted to a small range of items being bought.
  • It is about “age-groups”, not “generations” or “cohorts”.
  • There is no real “intergenerational” content.
  • It identifies recent trends.

So a more accurate title would be: “Some recent trends in spending behaviour by different age groups”.

Ways of comparing generations

I was working on this section before I knew about this report. I’m publishing it here because this report is mostly an example of the first way of comparing generations below, not the second way. (However, the report also adds “trends over about 10 years”, and these diagrams do not deal with trends).

A rather uninformative way

It is easiest and tempting (especially for younger people observing older people) to compare different generations at the same real date, for example 2013:

The obvious problem is that people in the 2 generations are at different life stages, and however fair the system is the comparison will show massive differences. That is life! Trends will be different for different generations if only because some changes matter only to one generation and some only to the other. (A change to the cost of end-of-life care only affects Generation P while a change to the cost of education only affects Generation Q).

Perhaps the main valid use for this sort of comparison is for marketing purposes, deciding what products to make and how to promote them.

Like-for-like comparisons

Comparing different generations at the same life stages may defuse some resentment. For example, comparing the age 15-25 stage of people born 1945-1950 and people born 1980-1985 will show startling differences. About 75% of people born 1945-1950 left school at 15, another 20% finished full-time education by 18, and only 5% went to university. This contrasts with a minimum school leaving age of people born 1980-1985 of 16 and 35% going to university. It puts so-called “free university” for the 1945-1950 cohort into perspective.

But while this may defuse some resentment, the differences are likely to be so wide-ranging and dramatic that few lessons can be learned. And how many young people really care about the early struggles of people decades older than them? They have their own current problems, and may lash out verbally in frustration and anger anyway.

Comments on “Spending Power Across the Generations” by the Intergenerational Foundation

The Intergenerational Foundation

These are comments on “Spending Power Across the Generations” by David Kingman, published on 23 January 2013 by the Intergenerational Foundation.

I believe for the reasons below that the report doesn’t make its case. Whether or not any of its assertions and conclusions are true needs information and analysis not included in the report and which may not even exist.

Generations versus age groups

There are 2 main uses of “generation“: familial (biological) generation, (not applicable in this report), and social generation or generational cohort. While the latter is often used in the context of intergenerational issues, for example “baby boom generation or cohort”, “Generation X”, “Generation Y”, “Millennial Generation”, that meaning is not being used in this report.

This report talks of people mostly in terms of age ranges, for example “pensioners”, “under-­30s”, “working-­age people”, “young people”, “People aged 50–74″, “people aged 65 and over”, “people aged 16–34″, “people aged 55-64 and 65+”, “all of the age groups below 45″, “People aged 65–74″, and “the over-75s”. (I spotted 2 generational cohorts: “the generation born before World War Two” and “baby boomers”).

This is not a trivial point, because over time an “age group” behaves very differently from a “cohort”. The individuals in a particular cohort stay the same (except for dying), while the individuals in an age group are a fluid set. This can have important, and sometimes counter-intuitive, consequences:

  • Example 1:
    The average wealth of an age group may decrease year by year even though every relevant individual may become wealthier year by year. (The “20-30″ age group may become poorer on average even if every individual passing through that age range is steadily becoming wealthier).
  • Example 2:
    The average wealth of an age group may increase even though the wealth of every relevant living individual may be constant. (“Pensioners” may on average become wealthier year by year even though all existing and new pensioners may individually have constant wealth).
  • Example 3:
    The proportion of an age group in the electorate may increase even though the total electorate is increasing. (Whereas a cohort relentlessly decreases as a proportion of the electorate once everyone in it can vote).

One consequence of the use of age groups instead of (generational) cohorts is that statistics about the age group over time do not necessarily reflect the life-experiences of the individuals passing through that age range. This is important for this report because it shows trends over about 10 years.

(An obvious consequence of having such a large variety of age ranges in the report is that the individuals being talked about in one paragraph are often not the ones being talked about in the next! The theme of the report is vaguely “young = unfortunate” and “old = fortunate”, where “young” and “old” are ill-defined).

Spending decisions

The report discusses spending on just 4 items: “Theatres and Cinema Tickets”, “Overseas Travel”, “Food and Eating Out”, and “Driving and Petrol”. The assumption appears to be that a fall in expenditure in these is caused by reduced disposable income while a rise is caused by increased disposable income. Like so many arguments used in intergenerational analysis, this is an obvious example of the cum hoc ergo propter hoc and/or post hoc ergo propter hoc fallacies.

A personal example: throughout much of my career I was something of a workaholic. Despite (then) being a highish-rate tax payer, I spent little on those items, partly because I simply didn’t want to spend my time on them. (I didn’t learn to drive until my early 30s so the “driving and petrol” item wasn’t directly relevant. I bought my first car out of income. I bought my latest car at 63 out of life savings). Now that I’m 65 and retired I still spend little of my income on them, preferring to spend my time and money on such things as house improvements and my main hobby (photography).

The inclusion of these 4 items says more about the interests of the author of the report than about society as a whole! They are not all the “important categories” he claims. Obviously “food” matters, but not “dining out”. Surely “savings and pensions” have to be included? Clothing and footwear, recreation and sports, household goods and furniture? And gas, water, and electricity, all of which have a discretionary element. “Transport” would be a better item than anything implying possession of a motor car, and to be fair the report does include public transport. And the magnitudes of this spending and these trends matters a lot, for example:

Roy Vice tweet

There are various reasons why spending on these items varies from person to person and/or over time. Spending on these items proves little about “spending power“, despite assertions, but mostly about “some spending decisions“.

(Note: I’m not denying the possibility that there is a causal relationship between actual spending power and these spending decisions; simply that the report itself doesn’t make the case).

The report concludes “This analysis supports the view that the older generation have enjoyed more favourable economic conditions during the first part of the 21st century”. I wish! It has economically been the worst decade of my adult life. But now that 38 years of mortgage payments are over, and I’m willing to spend my life-savings, and I’m getting my state pension, I’m comfortable. Appearances can be deceptive.

Using the “checklist for intergenerational unfairness”

Although the report is not explicitly stated to be about intergenerational unfairness it is likely to be read that way. I’ve used “A checklist for Intergenerational Unfairness” to see whether it can plausibly be read as evidence of intergenerational unfairness.

1. Is there more than one generation?

There are lots of age groups rather than generations, but I won’t quibble. The report has something to say relating to people of different ages.

2. Is it intergenerational?

Is it concerned with “relationships between children, youth, adults and seniors, particularly in terms of treatment and interactions“? Is there an element of cause and effect?

No. I don’t think the word “intergenerational” should be used about this report. (There is only one insignificant use within the text: “Restaurant bills also proved to be weighted intergenerationally”. But the report is published by the Intergenerational Foundation which is cited throughout it).

3. What is fair?

The report doesn’t identify by analysis what fair numbers would be. Therefore it cannot be said whether the trends imply unfairness. They may just be things the author doesn’t like.

4. Have the people concerned been consulted?

Apparently not. Their spending was sampled, but not their reasons.


My comments on this report are similar in several ways to those in “Commentary on “Intergenerational Fairness Index” by the Intergenerational Foundation“. This probably arises from a failure to start with a properly formulated hypothesis, then, in addition to identifying what supporting evidence is needed, to identify what evidence is needed to refute alternatives. This failure makes the process more like theology or ideology than science.

It is hard to deduce a person’s spending power from the narrow set of items covered by this report. There are too many other things that people may or may not spend their money on. (I don’t use a mobile phone except for emergencies, or have a smart phone or an iPad or mobile computing of any sort, in spite of being a webmaster for several websites, blogs, and social media accounts). Priorities change over generations.

The large variety of age ranges covered in the report gives the impression of having to grab whatever data can be found, rather than using data specifically to test the hypothesis.

I believe this report is really about:
“Some recent trends in spending behaviour by different age groups”.

Further reading

 Leave a Reply

Name required. (Spam NEVER gets published before being trashed)

Email required. (Spam NEVER gets published before being trashed)

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>