Some measures to stimulate “economic growth” don’t address underlying economic strengths and weaknesses. Instead they are skewed in order to manipulate – or “game” – the GDP indicator. The focus on GDP becomes a perverse incentive for short term bubble growth that benefits rent-seekers. Public debate on economic growth should focus on the underlying aspects of growth (ie. jobs, education, production) instead of doctoring the overarching measurement.
Economic growth is widely seen as a necessity for improving society, and GDP (Gross Domestic Product) growth is taken as the main indicator for measuring economic growth. This article will focus on the abuse of GDP, NGDP (Nominal GDP), and GNP (Gross National Product) as a perverse indicator and argue that it is not a useful as an indicator for economic growth beyond theoretical academic purposes. Indeed, when GDP growth is used as the measure of economic growth it transforms from being “just an indicator” to a goal. The goal of economic policy becomes an increase in GDP growth, as a proxy for “economic growth”, or even “economic development”. As a result, governments game the indicator, taking measures that lead to the GDP number to grow, without actual sustainable economic improvement. In this way GDP creates a perverse incentive to develop policies that lead to short term GDP growth, while undermining the long term perspectives for the economy and sustainable welfare.
Is gaming of the GDP a policy choice, or just driven by perverse incentives?
1. Indicators and goals
1.1. Briefings and media releases concerning the economic forecasts
I just read the economic forecasts for South-East Asia over the next few years from some major banks and multilateral institutions.
The organizations responsible for these documents have done a deep analysis, but the media and mainstream publications do not consider the details of these reports. The takeaways for the public are a focus on GDP it is forecast to go up by extrapolating from current GDP growth, adjusted to longer term trends.. The vital underlying economic analysis is deemed less important than the uphill tick of the GDP fetish.
Little media attention goes to whether China is getting out of the crisis or taking policy measures that make it worse in the long run. No comments are made on the reliability of the statistics for China. Little is said about how the 2 major growth countries in the region – Vietnam and the Philippines – increasingly build their GDP growth on remittances. When the GDP rises because of the remittances, it can lead to complacency, as everything seems fine. .
All eyes are fixed on the GDP numbers. As trust in a country depends on the expectations of the business community in that country, much is at stake. If GDP grows, business confidence goes up, leading to more Foreign Direct Investment, better conditions for trade, cheaper loans, etc.
A government is represented by its GDP grade in the world. The GDP is the indicator, the report card of a government and a country.
1.2. Perverse incentives: driving change in unexpected ways
Indicators work. Well chosen indicators are useful to get results . Measuring the number of children vaccinated against polio leads to fewer cases of polio, and eventually to the eradication of the disease. However, statistics can be falsified. Nurses can lie about the numbers of children vaccinated and doctors can fail to report cases. In the health system, the collection and use of quality statistics is of capital importance.
A perverse incentive is an incentive that has an unintended and undesirable result which is contrary to the interests of the incentive makers. The targets set on indicators used for measuring performance are easier reached by “gaming” the indicator than by doing what was intended.
Performance indicators in the banking sector rewarded risky behaviour by linking risky behavior to pay. Selling mortgages to people who could not afford them looked good on paper, but it led to the banking crisis of 2008. Performance measurement in the private sector seems so often gamed by management and staff, that HR experts question its usefulness.
More often than not, performance indicators are invitations to cheat, by employees but equally by companies and institutions, especially when financial consequences are attached to the Key Performance Indicator
Do performance indicators in the economy have the same effect? Could they be gamed?
1.3. Goodhart’s Law: When indicators become goals
Goodhart’s law is named after the economist who originated it, Charles Goodhart. Its most popular formulation is: “When a measure becomes a target, it ceases to be a good measure.”
The original formulation by Goodhart, a former advisor to the Bank of England and Emeritus Professor at the London School of Economics, is this:
“As soon as the government attempts to regulate any particular set of financial assets, these become unreliable as indicators of economic trends.”
Goodhart’s Law is specifically written for economic policy making: an indicator used for steering policy stops to be useful for steering the policy built on it, as people find ways to aim for the indicator without any positive effect for the underlying goal.
A very good take on Goodhart’s Law and GDP can be found on Phil Ebersole’s blog:
The aim is evidence-based policy. The result is policy-based evidence.
1.4. Aggregate indicators or disaggregated data
GDP is a typical aggregate indicator. It brings together information from all sectors, all layers of the population, all markets in an economy. This means that for policy purposes, the number is useless. Indeed, the different movements in each layer can mean different things: cutting 100 year old trees is GDP growth, taking a reusable bag to the shop shrinks it. Policy will seldom be general, so the underlying data should be used instead of the general number. However, if the aggregate indicator becomes the target itself, policy will be directed to move the general number. It will be aimed at the sectors in order to move the aggregate indicator, not specifically to have a sound policy for the specific sector. An aggregate indicator strengthens the effect of Goodharts’ law.
Due to the complex nature of economic growth it is impossible to attribute exactly the cause and effect of policy measure in the short-term. The effect on GDP is easier to measure. One of the underlying problems is the use of a composite indicator: Using one parameter with many degrees of freedom is not conducive to an informed public, policy, learning and steering. It leads to choosing the policy with maximum effect on the indicator rather than on economic development, which should be the real goal. Using disaggregated data on aspects of economic growth is more useful for measuring economic growth, and leads to less perverse policy incentives.
Confronted with the inadequacy of GDP to catch sustainable economic growth, former French President Sarkozy created a commission to study the question of indicators of economic growth. The report of this Commission on the Measurement of Economic Performance and Social Progress, the “Stiglitz commission”, proposed to ditch the GDP and replace it with a dashboard approach, with a set of indicators instead of just one number.
2. The GDP as a goal
2.1. GDP is a black box economical model
GDP is an economical model on its own. The model takes the different data from different sources, and spits out one number. The higher the number the better. It is based on a lot of assumptions and estimations. These assumptions give weight to some activities, and not to others, they are judgment calls. Estimating and including the rising quality of computers gives a seal of approval to digital progres but other improvements – like better schools results – are not included. So, when GDP is used as a target, the components that lead to the outcome are hidden. A growing GDP does not advertise the underlying assumptions and the improvement or deterioration of the non-measured aspects of the economy, nor whether this growth was sustainable or not. Growth coming from investment in residential housing, might mean the prices are driven up by speculative markets, or might mean that valuable additional houses are being build. In fact, for little elastic markets, adding infrastructure might lead to lower total market value, lower GDP when selling and buying.
As the GDP number hides the underlying realities for everybody except the data crunching experts, it acts as a black box. The public does not really knows what caused the growth, except maybe after a crash, with hindsight. This makes it very tempting for policymakers to game the indicator and introduce policies that promote specific interests and balloon the GDP while not contributing to a sustainable economy nor social welfare.
2.2. Perverse incentives for gaming the GDP indicator
Geoff Edwards, in “Political Arithmetick: Problems with GDP as an indicator of economic progress” concludes:
GDP does not detect symptoms of fundamental economic malaise such as an unbalanced penetration of imports into strategic industries; the sinking of household and public savings into economically terminal consumption rather than infrastructure and asset regeneration; starvation of research or sunrise industries; rising defensive expenditure on remediating pollution or social decay; rampant speculation on asset prices such as real estate or the stock market; the accumulation of private or public debt; or the transfer of assets to foreign investors.
He concludes: “By misleading those responsible for public policy about the nature of desirable economic activity, GDP drives perverse economic policies throughout the industrialised and developing world.
By using GDP growth as the goal of economic policy, the policy makers have options. They can pursue policies that will lead to long-term sustainable growth, such as widening and deepening education. They can promote productive investment and infrastructure investment with long-term growth effects, such as the building of railways, ports or social housing. All these policies are known to promote GDP growth in the long-term, probably with the first effects during the legislature of the successor.
In order to grow GDP fast, and getting re-elected, other measures work faster, while benefitting only a subgroup of the society, a subgroup with a lot of lobbying power: stimulating a real estate bubble or a stock market bubble; giving a free rein to extractive industries, leading to the Dutch disease and less competitive industries.
The real estate market is a prime example: residential investment alone is generally around 5 % of GDP, while housing services average around 12 to 13 %. This is a 6th of total GDP. During the bubble years (2005-2006) the residential fixed investment in housing was 6% of the GDP, During the crisis (2008-2009) this fell back to 3-3.5 %. When “normal” growth of the economy is around 1.5-2.5 %, the role of the bubble in the growth of GDP is important.
The crisis of the seventies was about gaming GDP on the demand side. The crisis of 2008 was about gaming the supply side:
- The post war governments created one of the major post war crisis by applying increasingly the Keynesian stimulus to generate growth during recession and boom. The initial stimulus in creating valuable infrastructure, was escalated to ever less efficient deficit spending. The infrastructure stimulus raised the GDP, created employment, and made the developers filthy rich.
- The neo-liberal supply side economics since the Reagan years escalated in gaming the GDP with ever more irresponsible banking deregulation and risk appetite, until the crash followed. “Only the bonuses were real money”
There are clearly options for “gaming” the GDP instead of taking measures for sustainable economic growth. And it is clear that some groups have an interest to push for this kind of measures as it benefits them directly.
It is more difficult to prove whether this gaming is a policy choice, or just driven by the perverse incentives
3. Why bad behaviour is almost always good politics*: The hijacking of the policy agenda through GDP growth.
The measures that form the basis for long-term economic growth are known: efficient and maintained infrastructure, generalised education of good quality up to university level, rule of law and antitrust legislation, an inclusive society, so all talents can be used in the economy and consumption does not dry up.
In a time when interest rates are negative, investments in infrastructure, maintenance, education, all of these investments have a very high return on capital.
This is not the policy we see around us. Instead we see austerity measures.
On April 15th, I read two blogs, One by Brad DeLong, the other by Tyler Cowen. Both refer to the gaming of the GDP by the elites for short-term rent seeking:
Brad De Long:
One way of looking at it is that two things went wrong in 2008-9:
Asset prices collapsed.
And so spending collapsed and unemployment rose.
The collapse in asset prices impoverished the plutocracy. The collapse in spending and the rise in unemployment impoverished the working class. Central banks responded by reducing interest rates. That restored asset prices, so making the plutocracy whole. But while that helped, that did not do enough to restore the working class.
Then the plutocracy had a complaint: although their asset values and their wealth had been restored, the return on their assets and so their incomes had not been. And so they called for austerity: cut government spending so that governments can then cut our taxes and so restore our incomes as well as our wealth.
But, of course, cutting government spending further impoverished the working class, and put still more downward pressure on the Wicksellian neutral interest rate r* consistent with full employment and potential output.
And here we sit.”
The simplest China model for 2016 is this. Due to the prevalence of SOEs and state influence in the economy, the country can in fact (for now) achieve almost any gdp target it wishes, at least within reason. But it trades off the quantity of gdp for the quality of gdp, and this time — again — the Party opted for the relatively high growth figure. That is bad news, not good news.
GDP growth is apparently being gamed as a matter of course. It has become a silly goal and a bad indicator for real life economic growth.
From the examples it seems some groups are proposing the short-sighted policies for gaming the GDP on purpose, because they benefit from it directly. This gaming is facilitated by the fact that the GDP-indicator contains perverse incentives for the policy makers to choose the bad policy, leading to easy, instant GDP rise, inequality and bubbles, above the better policy, which would require a real political and economical strategy and would lead to rising economic welfare, but only in the long-term.
Does it really matter if it is a conspiracy?
The policy makers and the elites can claim to the public and themselves that they are doing it for the good of all, and they can prove it: The GDP is rising.