Didn’t They Do Well! Is it Not Time We Got Over The ‘Generation Game’?
We have all used them, we all doubt them, but we all keep using them anyway even though we know they probably don’t quite represent reality. Like management fads and repeatedly pushing the elevator button we just can’t resist using ‘generations’ to describe how different groups of people think and act in life.
You can’t bake a cake nowadays without somebody advising you to account for generational differences. It usually goes something like this:
To bake a cake just mix butter, flour, sugar and an egg then pop it in the oven. When cool serve it up in single slices. Make sure you serve it in way that is appropriate for people’s generation – Baby Boomers will want to eat it with a fork leaving their children to clean up any mess they’ve left behind. Gen X will be too preoccupied with mortgage payments and career burnout to notice you were serving cake in the first place which is a good thing as ‘studies’ have shown that the gluten will probably kill then anyway. Gen Y will feel entitled to an ‘individualised’ slice but will inevitably resent the fact that their slice is smaller than what they had been led to expect. No need to supply utensils to Millennials as they prefer to scoop it up with their smart phones.
If you are in the benefits or pensions space, you probably have all sorts of people approaching you with ‘generation’ theories and how you must use them. I’ll raise my hand in guilt here as well. The labels are a simple shorthand and just too widely used to be easily abandoned.
However, I just can’t bring myself to use them anymore as are an increasing number of other people. For example, McDonald’s Chef Exec Steve Easterbrook speaking in relation to the challenges his business is facing said that they would like to see “… less simple talk of Millennials as though they are one simple group with shared attitudes.”
Mr Easterbrook is right – if you think even slightly deeply about generation analysis the whole thing starts to quickly fall apart. We can’t even agree on what the definitions of each generation are. Some people say Baby Boomers begin in 1945, others 1955. Just what is a Gen X’er anyway? Some say they are people born in the mid/late sixties – others much later. I remember when the term referred to young slacker types in the early 90’s but is now applied to anybody too old to be young but too young to be old. Turns out the term was originally coined by Robert Cappa (war photographer and co-founder of Magnum Photos) and referred to kids growing up during WWII. Now it’s being applied to the children of that Gen X’s younger siblings. Who’s making this stuff up? Is there an international standards agency that defines them? How can we say something is important to benefit strategy when we can’t even define it in any kind of reliable way?
It gets worse when you look at actual data – at least in the benefits space. When you examine what people are actually doing and why, the whole ‘generation’ thing stops looking like a good explanation of behaviour and preferences – or least not a good enough explanation. At the end of the day it’s the socioeconomics around income and family priorities that drives benefit decisions. In fact, I’d put age third in the pecking order after income and dependents in terms of explaining benefit selection behaviour.
We took eight common benefits and examined how four factors – generation, income, dependants and marital status – correlate with benefit take up rates where the employee has a choice of benefits and cash (see table below). We then ranked the correlations so rank = 1 means that the factor has the highest correlation with take up rates for that benefit, 4 the lowest.
Income and/or dependents are significantly more correlated with take up rates for most benefits except gadget loans and restaurant vouchers and to some extent health screening. We then looked at a mix of 20 benefits and found similar results with income and dependents being much more correlated than generation. It should also be noted that when you ignore ranking and look at the absolute value of the correlations, generation is only slightly more correlated with benefit take up than marital status. Again, we are not saying that generation (or really age) is a poor predictor – just that the data strongly suggests that it is not a primary correlate of benefits choices – more like a tertiary one that is much more useful when combined with income and dependents – something that generation analysis does not account for very well considering the range of income diversity there is in the UK and how the presence of dependents overrides youthful tendencies.
Stats and facts aside, I’ll give you an even better reason to avoid playing the generation game. My Baby Boomer father – retired and well into his 70’s – hates it when I call him on his landline. “Just call me on my iPhone!” he says, but I call him on his landline anyway. I don’t know why I prefer his landline (I never use my home landline). I think it’s a subconscious bias that drives me to use his landline which gets to the crux of why we want to use ‘generations’: they fulfil our prejudices about people.
We want to believe that people of different so called generations behave and have needs in stereotypical ways. But, that’s not reality as Mr Easterbrook has pointed out. Reality is that single people act like single people – Gen X or Gen Y because they have similar life pressures and priorities that are different from married people with kids. By the same token a Gen Y with kids will act just like a Gen X with kids because people with kids are family focused. People who want to use their smart phone for benefits want to do so because it’s more efficient – not because of their ‘generation’.
So there are two good reasons to avoid or at least downplay the generation thing: First, the data does not really back it up well enough when it comes to benefit choices. Socioeconomics around income and family are much better indicators of benefit preferences. Age does factor in; it’s just not really the whole story – more of a sub plot with exceptions at the extremes. The biggest age effects are the very young and very old – but driven by economic and physical realities: the young have lower incomes and fewer interpersonal responsibilities while the older have retirement related priorities and more health concerns. In other words, it’s the factors that correlate with age, not age itself that is driving at lot of benefits preference behaviour.
Secondly, but just as important, ‘generations’ or other labels placed on age groups just feed our prejudices and when you want to make smart business decisions impacting real people and involving real money, that’s not good. Making the right decisions boils down to using facts, evidence and avoiding unnecessary bias.
I am not saying ignore age – just that life is more complex than age based factors alone but not intractably so. Look again at the table above. It is economics and family, then age when it comes to benefit preferences.
What’s the alternative? Use age bands as they are more neutral and are a direct measure that can be combined with other factors. If you need to label them think in terms of early career, mid-career, nearing retirement or even combine age bands with service year estimates as both have real meaning in the context of benefits and pensions. But at the end of the day, income, dependents and marital status are big factors impacting savings and benefit behaviour, the former two being the primary ones and age bands or work/life cycle should be used in combination with these.
Let your employees tell you their own story through their data – you can measure benefit take rates and preferences as a function of socioeconomic factors such as those above (as well as age). There is no need have that story superimposed on your employees via potentially biased concepts of ‘generations’.
So take care with generational analysis and don’t refrain from challenging it – it’s a path laid with good intentions but there are better ways of capturing age effects.