I remember my first lesson in A-level economics primarily for a joke my teacher told the class. It went something like this:
Three men — a physicist, a chemist and an economist — became shipwrecked and ended up scrambling into a lifeboat until they came to a desert island. They had plenty of food, since the lifeboat was stocked with an abundant supply of canned food. Unfortunately, whoever packed the tins forgot to add a tin opener, and so the three men found themselves with enough food to last for over a month, but with no way of getting at it.
“Well,” said the physicist after giving it some thought. “I think the answer is obvious. All we need to do is to smack the tins on a rock and that way they should eventually open. Do you agree?”
“No,” said the economist, shaking his head disapprovingly. “I have a better idea.”
“Well?” said the physicist and chemist looking at him eagerly.
“How about we just assume we have a tin opener?”
No, I didn’t find it particularly funny either. But it does illustrate the point that economists can sometimes make assumptions which seem to skirt around somewhere in the suburbs of outlandishness.
Now of course all economists make assumptions, but reasonable economics must be based on reasonable assumptions. This is especially true in the area of Labour Market Information (LMI). One of the major areas of LMI analysis where people often make assumptions is by looking at data for a large geographical area, and then assume that the trends it reveals tell us what is happening in a smaller geographical area within that larger region. But is this a reasonable assumption to make?
Let’s take an example. Looking at the period 2008-2013, our data shows that the fastest growing occupation in Britain, as a percentage, was Dental nurses, which grew by 39%, from 42,138 jobs in 2008 to 58,436 in 2013. Now let’s say that back in 2008 we had looked at Dental nurses and had correctly predicted growth of almost 40% throughout the country over the following five years. Would it have been reasonable to assume that this growth was going to be consistent across the country?
According to our data, the growth was by no means uniform throughout the regions, with Yorkshire and the Humber, for example, seeing huge growth of 104%, whereas the likes of the North East actually saw a decline of 12% over the period. In other words, taking a national growth rate and assuming that it can be used to map out regional trends would be an unreasonable assumption to make since it could wildly underestimate growth in one region and wildly overestimate growth in another.
Now you might say that making assumptions about regions using only national data is clearly not a good idea, since regional labour markets can vary dramatically, but surely it would be reasonable to begin with a region and make assumptions about smaller geographies within that region? Actually no. Taking one region — Yorkshire and the Humber for instance — and drilling down to the county/unitary authority level highlights exactly the same issue that we had in the bigger sample. Over the period 2008-2013, the region as a whole showed growth of 104% in Dental nurses, but the county/unitary authority levels within the region shows wildly differing results. Kingston-upon-Hull, for instance, saw growth of 188% over the period 2008-2013, whereas growth in North and North East Lincolnshire was limited to 7%.
And the problem doesn’t stop there. Drilling down even further can often show the same issue again. Looking within the North and North East Lincolnshire county/unitary authority level to the local authority level again shows that growth at the higher level is by no means necessarily indicative of the lower level trends:
|Local Authority||2008 Jobs||2013 Jobs||Change||% Change|
|Source: EMSI Covered Employment - 2014.1|
|North East Lincolnshire (00FC)||85||197||112||132%|
|North Lincolnshire (00FD)||218||126||-92||-42%|
The historical data above reveals the following truth: what goes for the national economy does not necessarily go for the regional economy; what goes for the regional economy does not necessarily go for the county/unitary level economy; and what goes for the county/unitary level does not necessarily go for the local authority level economy. In other words, it would be extremely unwise to base our assumptions of what is going on at the lowest geographic levels on what is going on in the higher levels. Economies don’t always work in this way.
This is why we labour to source our data from a variety of datasets, then spend much time and effort modelling them together, binning the weaknesses and retaining the strengths of each dataset, to give a complete picture of the economy, from the national level right down to the local authority level. Taking data that only goes down to regional level and assuming that it can tell us what is going on at the lower level geographies will almost certainly lead to wrong conclusions. And if we are brought to the wrong conclusions about the data, we will almost certainly find ourselves on the wrong track if we use that data in our strategic planning, potentially leading to all sorts of wrong decisions. Unreasonable assumptions can sometimes turn out to be dangerous assumptions.
For more information on our data, and our approach to Labour Market Information, contact Andy Durman on firstname.lastname@example.org