The Limits of GDP

“Distinctions must be kept in mind between quantity and quality of growth, between costs and returns, and between the short and long run. Goals for more growth should specify more growth of what and for what.” – Simon Kuznets, developer of the measurement of GDP on the limitations of the statistic.

There’s a measurement of technological advancement in science and science fiction circles known as the Kardashev Scale. It measures how advanced a civilisation is by how much energy it consumes or is able to harness using its technology.

A Type 1 civilisation is able to use the same amount of energy as falls on their home planet from their home star. For Earth-Sol, this is a power consumption of about 7×1017 Watts.

A Type 2 civilisation is able to harness the entire energy output of its home star – either by colonising many star systems or by fully enclosing their home star by energy gathering technology like a Dyson Sphere. For our star, this is about 4×1026 Watts or about 500 million times the energy consumption of a Type 1 civilisation.

Finally, a Type 3 civilisation has the technology to harness the entire energy output of a galaxy – a feat rarely reached in even the most ambitiously scoped sci-fi but see Stephen Baxter’s Xeelee Sequence for beings which are capable of at least this and more. This level of tech involves the consumption of about 4×1037 Watts or about 100 billion times the power used by a Type 2 civilisation.

And where does our wee planet sit in all of this? Carl Sagan extrapolated from these points to create a more comprehensive mathematical formula for calculating a given civilisation’s Kardashev number. He also estimated in 1973 that humans generated an average of 8 Terawatts of power which gave us a Kardashev Number of about 0.70. As of 2015, our energy generation has increased to an average of about 19 TW which makes us now a Type 0.73 civilisation. Extrapolating out linearly, we should hit Type 1 around the year 2325.


Imagine that the only thing that the scientific community really cared about was our progress up this scale. All of science was bent towards our expanding our power generation and consumption base. And all that politicians cared about was their “Kardashev Growth per year”. It wouldn’t matter if your discipline was physics, chemistry, biology or whatever. Anything that grew the Kardashev number was funded. What would we be missing if that was all we cared about?

(For my academic friends who want to draw parallels with the current obsession with citation number and impact factors…you’re not wrong, but…next time.)

It’s worth noting that whilst total global power generation has increased by a factor of 2.4 since 1973, global population has only increased by a factor of about 1.8. The “global average human” today is using about 33% more power than the “global average human” of 1973. This is due to a massive reduction in global poverty and an unbelievable expansion of buying power and access to technology around the globe.

Of course, there are limits to this kind of measurement. Technology doesn’t just get more powerful and more widespread as it develops, it also gets more efficient. The fitness band I’m wearing right now weighs less than 100g but has orders of magnitude more computing power and uses orders of magnitude less electrical power than my first desktop PC from the mid-1990s – never mind the room-sized behemoths of the early days of computing in the 1950s and ’60s.

We also must think about the environmental impact of our technology – another major difference between 1973 and now is the exponential expansion of renewable energy technology at the expense of fossil fuels.

A 2015 Megawatt emits about 20% less carbon dioxide than a 1973 Megawatt and that figure is about to fall dramatically over the next decade as we electrify our heat and transport industry.


So obviously, rendering a civilisation’s level of technological development down into a single number and then measuring the growth of that number might be interesting and may be useful as a very broad comparator but it probably doesn’t tell anything useful about the shape of that society or what’s going on inside it.

So why do we expect the same from other renderings such as Gross Domestic Product?

This measure – essentially the sum value of all economic activity within a country in a year – was developed in its modern form in the 1930s and has since become the gold standard of national statistics. Huge swathes of statistics are measured either in relation to it (think of “national debt as a % of GDP”) or as a derivative of it (GDP growth” or “GDP per capita”, for instance).

But even at its inception, the creator of it noted the limitations. A measure of the growth of GDP says very little about how that growth has occurred or who has benefited from it.

I was able to discuss some of these issues when I visited the Scottish Parliament’s Economy, Jobs and Fair Work Committee earlier this week as well as in an article in CommonSpace yesterday.

The Scottish Government has, to its credit, tried to improve some of the measurements it uses via program like the National Planning Framework. One of these is the idea of “inclusive growth” which explicitly states that measures should be taken to ensure that when economic growth occurs, it should not simply accumulate with the top few percent.

It’s a laudable goal but I still have my reservations in that the metric is still pinned onto the concept of “growth”. Just as a growing economy may accumulate wealth in the top, it’s possible that a recession could increase inequality too. Imagine a drop in the economy that causes mass unemployment, reduction in working hours, an increase in casual, part time and “gig” labour whilst at the same time the wealthiest are bailed out by the state. You know…a bit like what happened in 2008.


Just as it’s possible to try to ensure inclusive growth, it should also be possible to ensure inclusive de-growth when recessions come and inclusive stasis when economies are growing. The government would be well advised to focus as much on inclusivity as it does on growth, maybe more so.

To be fair, there was at least some logic to using GDP as our measure of economic development. For many years, wage growth was pretty firmly linked to economic growth. If the company you worked for did well, you did well. But two great shifts have happened in the past 40 years. The first came in 1979 with the rise of neoliberalism. The economic programs enacted by Thatcher and others were coupled with a decoupling of wages and growth such that the latter started to pull away from the former. The gutting of the unions and the reduction of collective bargaining was perhaps chief among those reforms.


The second decoupling came in the wake of the 2008 Financial Crisis. This time it came in “productivity” – the amount of GDP generated by a given hour’s labour. It’s a complex topic but part of this has come from the rise of the “gig economy” where folk have – often through somewhat convoluted means – become “self-employed” and have lost access to even individual bargaining. With no commitment at all to the company that they “service” (rather than work for), it’s no surprise that innovation and investment have declined. Not that it’s the fault of those workers. With no commitment to them, the company needn’t care beyond short term profit extraction and share price either.


Another possible reason for the decoupling of productivity and wages (as well as the stagnation of productivity overall) is the rise of certain sectors of the economy that don’t quite fit with older measures such as “added value” and suchlike. I’m talking here about jobs like personal care. It could be possible to “increase productivity” in personal care but if that means “service more clients per shift” then what you’re asking carers to do is cut visiting times from 30 minutes down to 15 minutes down to 10 minutes…

So maybe we don’t want “productive” care. I’d prefer effective care. And maybe jobs like care don’t produce huge feedback multipliers in the economy but there are other ways that such care could be considered “valuable”. Especially to the person receiving that care.

So, just as the Kardashev Scale is for technology, GDP should only be dubiously taken as the be-all-and-end-all of economic development. Politicians really should start taking note of what’s going on underneath that headline number. It’s a harder story to tell, true, but economies are far too complex for the “simple” story to be sufficient.

And if it turns out we’re simply not measuring enough to answer our questions, well…that’s a story for another day.

TCG Logo 7


One thought on “The Limits of GDP

  1. Hi Craig, heading off at a tangent, I know, however I think this might interest you not so much on the track you are following, but in a paper that suggests a link between GDP and Bond issues.

    Sovereign GDP- Linked Bonds
    Published on 21 September 2016
    Financial Stability Paper No. 39
    By James Benford, Thomas Best and Mark Joy, Bank of England, with contributions from Mark Kruger, Bank of Canada, and the Research Department, Central Bank of Argentina

    While the idea of governments issuing financial instruments whose repayments are indexed to gross domestic product (GDP) is not new, the current global backdrop of high sovereign debt coupled with low interest rates and weak and uncertain nominal growth prospects suggests the case for doing so may be especially strong now. This paper discusses the pros and cons of GDP-linked bonds, looks at when it might be most beneficial to issue, how investors might benefit, and possible ways of addressing the first-mover problem. The aim of this paper is to stimulate debate rather than provide answers.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s