As a young child I was greatly confused by Lewis Carroll’s marvellous poem, The Walrus and the Carpenter. What puzzled me so much was simply this: What on earth has a walrus to do with a carpenter? Isn’t one a blubbery sea creature with tusks and a moustache, and the other someone who makes things out of wood? The connection somehow escaped me. A little later, it dawned on me that this was precisely the point. Lewis Carroll’s stock-in-trade was writing poems and stories that were essentially nonsense, bringing all sorts of weird, wonderful and utterly disparate characters onto the same stage.
Now, if I were to suggest a connection between such nonsense and Labour Market Information (LMI), you might think this was … well nonsense. Yet if the methodology used to turn raw data into LMI is flawed, we can actually end up with a set of results that essentially treat occupations that are vastly different from one other as if they were — economically speaking — much of a muchness. We might not quite get as far as joining a walrus and carpenter together, but unsound methodology can easily give us bizarre scenarios such as The Nursery Nurse and the Pest Control Officer or The Undertaker and the Teaching Assistant, treating these disparate occupations as if their rates of growth or decline are bound to be the same.
We covered the basic principles of this in this piece, but let’s just recap. Occupations are divided into four classifications (SOC codes) with the most generic occupation level — say, Skilled Trades Occupations — being given a 1-digit SOC code, and the most specific occupations — say Sheet metal workers — being given a 4-digit SOC code. The chart below shows this in more detail, starting with the generic 1-digit classification, Caring, Leisure and Other Service Occupations:
So we have Caring, Leisure and Other Service Occupations at the top, which then branches out into two sub-divisions: Leisure, Travel and Related Personal Service Occupations forming one subset, and Caring Personal Service Occupations forming the other. Beneath that, we have a subset of 3-digit occupations, followed by a further subset of 4-digit occupations (I have colour coded the 3 and 4-digit occupations to show the connection between the specific 4-digit occupations and the parent 3-digit occupations to which they belong).
What you will hopefully notice about this is that the 4-digit categories contain an array of jobs which are often vastly different one from another. For example, Pest control officers and Teaching assistants clearly have little in common, either in the nature of the job itself, or in their economic connectivity. Yet it is crucial to note that the 4-digit occupations in the left “bucket”, despite often being very different, all fall under the 2-digit Leisure, Travel and Related Personal Service Occupations, whilst the 4-digit occupations in the right “bucket” all fall under the 2-digit Caring, Personal Service Occupations.
Now what would happen if we used a methodology that only drilled down as far as the 2-digit level, and then simply assumed that the rate of change found at this level will be the same for the 3 or 4-digit level occupations beneath it? This would be tantamount to saying that Playworkers have the same rate of growth and change as Undertakers, mortuary and crematorium assistants. This is clearly the kind of claim that belongs in Lewis Carroll territory, and yet some LMI solutions take exactly this approach.
Here’s an example of what this looks like in practice. According to our data Leisure, Travel and Related Personal Service Occupations (2-digit SOC) grew by 2% across Britain from 2008-2013. So does this mean that all the 4-digit occupations beneath this parent category also grew at 2%, as some LMI methodologies assume? The following graph suggests otherwise:
If we were to use a methodology which mapped the rate of change at the 2-digit level straight onto the 4-digit occupations, we would conclude that the number of Sports and leisure assistants has grown by 2% over the past five years. Except they haven’t. In fact, the numbers actually dropped by 4%. Using the same methodology, the numbers of Housekeepers and related occupations apparently also grew by 2% over the past five years. Except they didn’t. In fact, the numbers actually rose by 7%.
Why is this important? If you are going to use data to make crucial decisions about future strategy, you need to make sure you know the real rates of change of specific occupations, rather than lumping vastly different occupations into the same bucket by assuming that the rate of change is the same as that at the more generic level. Failing to do this means that you could well end up planning your strategy on a false basis, believing that an occupation is set to grow by, say, 6% over the next few years, when in fact it is set to decline by 8%. This could be a costly mistake to make.
The Beautician and the Caretaker may well make some fine nonsense poetry. Whether you want to base your curriculum planning on it is another matter entirely.
For more information about our LMI or our methodology, contact Andy Durman at firstname.lastname@example.org.