First published by Common Weal
The Fraser of Allander Institute does a good job of modelling the Scottish economy and isn’t known for hype – so why do I instinctively not really trust their statistics on what at face value appears to be a sudden and pretty enormous increase in the size of Scotland’s renewables industry?
To explain this I need to explain a little bit about modelling, a bit about what economists mean by modelling, a bit about the impact of clients and a bit about the lack of a clear headcount. Basically too many people in Scotland have an interest in taking the most positive possible view of our renewables industry and the giant problems with economic modelling make it a bit too easy to justify that.
The reason this is important is because Scotland’s economic future isn’t a conceptual thing or a conglomeration of statistics but a crucial, real-world reality which cannot afford to arrive at a point 20 years from now to discover that all of the ‘supported jobs’ in renewables turn out to be newsagents and marketing men (I’ll explain in a moment) and we’ve failed again to build a new industry.
You know, precisely like happened over the last 20 years.
First, we need to understand what economic modelling is because it has been so overwhelmingly abused in recent decades that we take for granted statistics we don’t really understand and which don’t tell us what they appear to.
The standard economic model was created in the US in the 1970s, and its what’s known as an input-output table. This is like a complex formula with many thousands of individual assumptions about what input leads to what output. These are not neutral, so for example if you write an I/O model which assumes that tax cuts will increase economic growth, every tax cut you feed into the model will be biased towards showing that tax cutting boosts the economy.
The I/O model we use of course contains that assumption (though some other assumptions do work in the other direction). And this broad model has endured since, largely unchanged. In fact largely not understood – few people have actually dived into the thousands of assumptions that make up this table to assess them.
The point here is that if you reverse the assumptions in even a small proportion of the elements of an I/O table, it can give exactly the opposite result. I’m not claiming this is crooked, just that it’s not neutral and its purpose is not what you think it is.
Because you probably think that these economic models ‘predict’ what will happen, and they do no such thing. They compare what would happen based on two different actions. They are meant to indicate ‘broad direction of travel’, not exact outcomes.
For example, if you reduce spending on transport, transport infrastructure deteriorates and this has a knock-on effect across the economy. Inputting cuts to transport spending will therefore produce a result with negative effects (and, because of the assumptions, some positive ones as the money is spent somewhere else).
Which means the model isn’t telling you what will happen but rather what difference might occur between two possible paths forward. Not ‘if I cut transport the following will definitely happen’ but ‘if I do or don’t cut transport then this model suggests the difference in outcomes would be in the order of’. You’re trying to compare two imaginary outcomes to get a sense of what they could mean.
If the workers putting up a wind turbine buy a Daily Record to read the sport at lunchtime, the newsagent who sold the newspaper may well end up being counted in the ‘supply chain
The second part of the modelling process is the data input. You have to add a model of any given economy onto the I/O table to reflect that economy, so you need detailed statistics on the state of the Scottish economy. The outcome of your model will depend on the quality of your statistics.
And you know where I’m going with this; in the Fraser of Allander report it makes clear that it didn’t actually have properly-defined statistics to work with so had to ‘deduce’ those statistics from other statistics. There are clear question marks about how accurate this could be even as a comparative tool.
But none of this is static; it all depends on the next set of assumptions fed in by the modeller and influenced by the client. I can tell you about this first hand because I commissioned an economic impact assessment back when I was working with the universities from a well-known economic consultancy.
Suffice to say we ran the model twice, once to see what it would say with what we thought the right assumptions were and then once again with those assumptions tweaked to give us the outcome we wanted. This isn’t ‘illegitimate’ because this is how everyone does it. When you’ve got one set of assumptions (the approach to what you’re feeding into the model) on top of another (the data on the real Scottish economy) on top of another (a 50-year-old complex American formula), many outcomes become perfectly viable.
So I am certainly not suggesting that Fraser of Allander has done anything in any way inappropriate here, just that no-one screws over their client and the client here is the lobbying body for the renewables industry with a strong interest in presenting everything as ‘fine’. It is inevitable that the economists will steer down a legitimate path that gives a positive outcome.
And the key to understanding that is in the word ‘supported’ and the interpretation of ‘supply chain’. Let me be really clear about this; if the workers putting up a wind turbine buy a Daily Record to read the sport at lunchtime, the newsagent who sold the newspaper may well end up being counted in the ‘supply chain’ or will definitely be ‘a job from induced extra spending’ (supported).
Or have you ever been phoned by a telesales marketing person telling you that there are grants for heat pumps if you buy them from them? That person will count as a directly-employed part of our renewables workforce, despite it being just a unskilled call centre job.
Even those supplying actual components for renewables systems (a legitimate supply chain measure) could just be driving them there from the depot where a multinational not owned in Scotland bought the components from China. I suspect a large part of Scotland’s renewable supply chain is really marketing and retail.
Or let me give you another example of what this all means. If the Scottish Government was to spend its whole budget on chocolate you could easily demonstrate a significant economic boost. Someone has to transport it, someone has to store it, someone has to be a night watchmen. That’s all GDP.
If you think it seems strange that there was a genuinely massive one-year increase in the number of renewable jobs and that that increase occurred right in the middle of a pandemic, you’re probably right
The point is that an economic model is supposed to tell you what is different between two things, not what has got bigger or smaller (you have to subtract ‘no schools or police’ from ‘chocolate boosts the economy’). To measure what has got bigger or smaller you need to measure the thing – which means you have to go out and gather accurate statistics about the industry directly from observation of the industry.
The model would then tell you how much of this is additional because of renewables and how much of that is offsetting something which is already there. For example, you could say ‘we’ve turned off this much electricity generated with gas and replaced it with this much electricity generated from renewables – how many jobs were lost, and how many were gained?
Or think about this way – we don’t include the jobs from 30 years ago that were briefly created when a gas turbine generation plant was built but we are counting the jobs for the erection of wind turbines, jobs which will cease once the facility is built. In theory, if it takes 100 people to service a gas turbine but only 20 engineers to maintain wind turbines producing the same amount of electricity, you would have actually ‘destroyed’ 80 jobs.
I cannot emphasise enough that I’m not having a go at Fraser of Allander here, an organisation for which I have a lot of respect. What I’m trying to tell you that if you think it seems strange that there was a genuinely massive one-year increase in the number of renewable jobs and that that increase occurred right in the middle of a pandemic, you’re probably right.
Because it probably didn’t happen in the way that it sounds like it happened. There just wasn’t a massive jump in high-value manufacturing and engineering jobs in the renewable sector in 2021, a year we were mostly still all locked in our houses or emerging slowly, blinking into the light of normal life.
Unfortunately the corporate renewable energy industry wants you to think all is going well so it maintains its generous treatment from government (bloody ScotWind), the Scottish Government desperately wants any evidence that its renewables strategy is working brilliantly and the Fraser of Allander want to produce a report which is 100 per cent justifiable but keeps their client happy.
And every part of this is done at a desk in front of a spreadsheet. What no-one is doing is actually going out and mapping properly where the renewable energy jobs are and what the quality of those jobs actually is. If Scotland is ever to escape its serial failures over renewables, that’s the number we need, not an assumption built on an assumption built on an assumption built on a commercial relationship.