GROWTH COMMISSION RESPONSE: 2. STRETCHING THE EMPIRICAL DATA

The Growth Commission report is 354 pages long. The Summary takes up 41 pages and contains 224 distinct clauses (one alone of which contains 30 sub-clauses). It’s fair to say this report is not designed to be an easy read and that only the most committed drudges (like this author) will actually read it all.

Given this, it is informative to see how the report has been spun by its authors and sponsors, to see what they have tried to get people to take away from it. We can then see whether the report backs up the claims made on its behalf.

Ahead of the report’s publication a press release^{1Framework for Scotland to emulate world’s best performing economies} was issued which previewed the report’s “core finding” that “an independent Scotland can emulate the world’s 12 best performing small advanced economies (SAEs), closing the growth gap and driving GDP per head to the median of these best performing countries”.

In fact the report observes how those countries perform and asserts that Scotland should aspire to these targets – that’s not the same as showing that Scotland can emulate these countries. There is nothing wrong with a statement of ambition, but it should be recognised as such and not as something more.

More importantly, we should focus on the quantitative claims made in the press release because these are claims on which the entire economic case (insofar as one exists) is founded:

1. “The report concludes that achieving this would be worth an additional economic output in Scotland equivalent to an extra £4,100 per person in Scotland”
    
2. “The analysis shows that small economies have performed better than larger ones consistently by around 0.7 percentage points per year in economic growth rate, over the last 25 years on average."

It is worth noting that both of these figures are based entirely on observations about the aggregate performance of economies in the carefully selected cohort already discussed (so the use of the phrase ‘small advanced economies’ is misleading). There is no direct link at all between either of these figures and the actual policy recommendations made in the report – yet these figures are used not just to scale the aspirational goal but, later in the report, are also implicitly used to scope how long it might take to reach it.

Taking the first claim, this is merely a corollary of saying “if Scotland’s GDP/capita was the same as the Netherlands, we’d have £4,100 more GDP/capita”. It’s an observation that, while unimpeachable as a statement of the obvious, leaves something to be desired in terms of insight.

The analysis is no more sophisticated than ordering the 12 ‘peer group’ countries in order of descending GDP/Capita and comparing Scotland’s GDP/capita to that of the country in the middle (i.e. the ‘median’ country, in this case the Netherlands). Given that the peer group excludes countries with materially lower GDP/Capita than Scotland^{2New Zealand is the only country included with a lower GDP/Capita than Scotland, its GDP/capita being 2.7% lower – advanced economies excluded without explanation are: Slovak Republic, Israel, Portugal, Czech Republic, Greece}, it’s inevitable that Scotland sits below the median point. The arbitrary nature of this comparison cannot be stressed enough. We could argue for different cohort structures and end up with different median countries and therefore different headline “achieving this would be worth” figures, but this would be merely to indulge in statistical gerrymandering of our own.

The Growth Commission looks at the GDP/capita of the Netherlands and asserts “This provides a measure of what Scotland can achieve in the future if it has the same ability to tailor economic policy to its own needs and advantages as these other countries do."^3A1.32 There is an implied causal link between the observed GDP/capita difference and ‘having the same ability to tailor economic policy’ – but at no point does the report actually make a case for this being the main, let alone the sole, explanatory variable. Factors such as membership of a particular trading bloc (or not), natural resources, climate, geographic proximity to other markets, language, population density, political stability, cultural work-ethic and many others may be better explainers of GDP/Capita differences than just ‘being independent’.

Similar problems exist with the second claim, which is based on the mean average real GDP growth rates of the ‘peer group’ versus the larger advanced economies. The Large Advanced Economy (LAE) cohort has not been pre-selected for success, so immediately we can see that this isn’t a like-for-like comparison. There are any number of ways we could argue to re-cut these comparisons, all of which lead to a lower observed GDP growth gap over the last 25 years.

• Using the Growth Commissions definition of successful SAEs and comparing to their defined LAEs we get their growth rate difference figure of +0.65%^{4a trivial point, but the GC has rounded this figure up to 0.7%}
   
• If we apply the less biased (although still questionable) criteria of comparing all SAE’s with population greater than 3m but less than 10m against all LAEs with populations greater than 20m we get a growth rate difference figure of +0.31%. This change alone more than halves the apparent growth difference and therefore more than doubles any timeframes which rely on this assumption.
   
• Alternatively, given that the report asserts Scotland’s “aspiration for EU membership”^5B8.63 and that this would be likely to involve adopting the Euro^{6The report is equivocal about this, stating “Scotland would join the euro only if and when such a decision is in the best interests of both Scotland and the EU, and the relevant criteria of the Maastricht Treaty were met” [C4.15] – but given the currency challenges an independent Scotland would face and the EU’s stated criteria for new member states, this seems at the very least a realistic long-term scenario (if economic conditions could be met)}, we can look at the performance of ‘Eurozone SAEs’ against the LAEs.^{7We have stayed with the Growth Commission’s definition of LAE’s here, so high growth Taiwan is excluded (thereby favouring SAE’s in any comparison)} This shows a growth rate difference figure of +0.26% over the last 25 years (dropping to just +0.07% since 1990). This comparison is particularly relevant because being an EU member states places constraints on the economic strategies a country can pursue (state-aid restrictions and requirements to adhere to EU trade agreements being two obvious examples).
   
• Following the logic of the report’s actual recommendations, we could take the GC’s SAE cohort and simply exclude Singapore, Hong Kong, and Ireland^{8There is also an issue with Ireland’s GDP having an unusually high share of overseas ownership (largely because of the low tax model). GNI would arguable be a better measure, but unfortunately the IMF does not record this figure. See Ireland’s ‘de-globalised’ data calculate a smaller economy, FT, 18th July 2017}, because the Growth Commission explicitly rejects their low tax^{9“not competing as a low tax location” [3.58, and various other locations]; “unlikely that [Scotland] can successfully operate an Ireland/Singapore model” [A3.77]; “little sense in competing as a low cost or low tax location” [A3.82]} models.^{10The Growth Commission also rejects the Singapore and Hong Kong models based on their high income inequality outcomes [Figure 2-9]} Just taking these three countries out of the SAE cohort actually reverses the growth gap, with this adjusted SAE cohort in fact growing slightly more slowly than the LAEs, a difference of -0.11%.
   
• As a last test, we can look only at the performance of those SAEs that the Growth Commission goes on to suggest Scotland should most seek to emulate (Denmark, Finland and New Zealand)^{11“we are especially drawn to a hybrid of Denmark, Finland and New Zealand” [3.58]}. The average growth rate of these countries compared to the LAEs is immaterially different at just +0.06%.
   
• The report itself even observes that “several small Northern European economies have generated sluggish growth performance since 2000 (notably Denmark, Finland, and the Netherlands)”^12A2.6. This is undeniably true as all of these countries have in fact seen slower growth than the UK since 2000.^{13Source IMF: UK = 1.87%, Den = 1.14%, Fin = 1.49%, NL = 1.39%}

All of this shows us that the ‘+0.7% superior growth’ figure which emerges as a headline from the report is not a robust assessment of how small advanced economies in general fare against large ones.

In fact, there is no empirical evidence to suggest that following the models of those economies the Growth Commission most seeks to emulate will deliver greater GDP growth than Scotland could expect to achieve while remaining within the UK. This point is clearly shown by Figure 2:

As well as illustrating the greater growth volatility experienced by SAEs, this graph also shows that New Zealand is the only one of the ‘chosen three’ which shows reasonably consistently superior GDP growth to the UK.  But focusing on GDP growth alone masks an important issue which the report itself highlights: “New Zealand’s migration-driven growth model [means that] despite GDP growth of 3.6%, per capita income growth has been around 0.6%.”^{14A1.92; according to IMF data, New Zealand’s 2000 – 2016 population growth was 1.28% pa compared to the UK’s 0.68%} Give the Commission recommends migration-driven growth, this is a topic we’ll return to.

In the context of the actual policy recommendations the Growth Commission goes on to make, the +0.7% figure is frankly irrelevant. There can be no justification for using this figure when assessing the likely impact of the report’s actual recommendations.

In no way does this negate the value of seeking to learn from others’ success – but to have a constructive and honest debate we must be realistic when assessing the medium-term potential upside for Scotland’s stand-alone economy.

<< Previous Chapter             >> Next Chapter

Quick-links to all Chapters

Executive Summary

Context & Response

1. Smaller isn’t Necessarily Better

2. Stretching the Empirical Data

3. Failing to Make a Case

4. A Reality Check

5. The Truth about Austerity

6. Aiming Too Low

7. The Missing Model

8. Currency – an Unsolved Conundrum

9. Making the Case for Union

Endorsements