The Great Invention: The Story of GDP and the Making and Unmaking of the Modern World - Ehsan Masood (2016)
Epilogue. Unfinished Revolution
Most countries have more reliable statistics on their poultry and egg production than on their output of discoveries and inventions.
All editors of newspapers and magazines have their “golden rules.” These are designed partly to enable publications to be distinctive, to reflect the values they hold, and to reflect the values of their readers.
In the Research Professional family of publications, one such rule is that we don’t use the word “foreign” to describe people who happen not to be citizens of the country where they work. It makes no sense to us because research is a highly mobile profession. The majority of our readers are likely to find themselves working away from their country of citizenship at some point in their careers. Each of us will be foreign at some point in our lives.
Another of our golden rules is to help readers cut through the fog of corporate jargon found in official reports and press releases. Often this is just the result of lazy writing, but words can also be used deliberately to convey ideas that can be laden with baggage. Editorial colleagues are advised, for example, to think carefully before using the word “investment” in their articles, particularly when it comes to reporting stories about new government funding schemes.
This might seem a small, perhaps trivial point, but it isn’t. We need to be cautious because governments have become fond of using “investment” when they really mean “funding,” and the two are not the same.
Investment is more often associated with business (as we know from the GDP formula). When businesses invest, they do so because they expect a return on their investment; they expect to make more money than they put in. When governments, on the other hand, provide “funding,” they are doing so as part of their role to protect citizens and give them opportunities to prosper. When governments provide funding to upgrade a hospital, resurface potholes on a street, or open a new university lab, that isn’t an investment, because they expect no monetary return. Government funding is part of the state’s bargain to protect those who elect them and contribute their hard-earned money in the form of taxation.
Governments fund; businesses invest. Or so I thought.
In April 2013, in the midst of writing this book, a headline from the Financial Times told me that governments who declare science spending as “investments” did have a basis to use that word. I also learned that what I was reading was to have profound implications for how we construct GDP and for ongoing efforts to reform it.
I discovered that it is actually possible to change GDP.
The Financial Times story was on the inside pages of my morning read and the headline was simple enough: “Data Shift to Lift US Economy 3%.”1 The FT announced to the world that, more or less overnight, the size of the US economy had increased and that this was partly because of a change to how GDP is calculated.
That change involved a kind of arcane accounting trick: it meant moving the column that contains spending on science to a different place in the GDP table. Up until then, everything that businesses and governments spent on science had been regarded as an expense with no expectation of financial return, in the same way that governments spend on hospitals or schools. But now, according to the FT article, science spending would be moved to the “investment” column. The reason for this shift was because of a belief, increasingly common among some economists and more politicians, that spending on science makes money. That is certainly true of science spending by businesses, which in rich countries constitutes the majority of science spending. However, it is more controversial than that.2
This change to how GDP is worked out had better consequences for those nations with big science budgets and big multinational companies. Countries such as the United States, Britain, Germany, and France would see their GDP revised upward by between 1 and 3 percent.3 Countries with much more modest research spending and few if any multinational corporations (essentially most of the rest of the world) would see little or no increase.
One of the frustrations of writing this book has been seeing how the struggles of so many great talents had so little effect on changing GDP. Mahbub ul Haq probably came closest to creating an effective global alternative to GDP in the Human Development Index, but this had zero effect on GDP itself. Maurice Strong tried to circumvent the problem by creating an elaborate infrastructure of environmental checks and balances, which did more to raise awareness of GDP’s limitations but did not tackle its problems head-on. Meanwhile, Bhutan’s fourth king Wangchuck tried to steer his economy using Gross National Happiness alongside GDP. But none, it seems, had given due thought to the elephant in the room. Only Robert Costanza had showed the world what the numbers would look like if GDP could change to stand for better things. Only he was unable to show how to get there.
What few of us knew or understood is that the change to how GDP is calculated that the FT was reporting did not happen overnight. It was the culmination of a decades-long effort spearheaded largely by the developed countries and led by the their trading club, the OECD. That FT story confirmed something known to its practitioners, which is that GDP has been modified—we also know this from the published revisions to its methodology since the 1950s. But the fine-grained process of how this has happened, and, more important, the politics of how change was brought about, remains opaque to the general reader, as there is comparatively little scholarship and even less journalism around these issues.
That FT story gave me a window of opportunity to try to resolve that question. The rest of this chapter summarizes my attempts to discover how GDP can change.
The story of how the OECD member states, and especially those who were research-intensive economies, succeeded in making a change to the composition of GDP starts half a century earlier and involves two of the characters we’ve already met in The Great Invention.
One was Alexander King, the urbane British civil servant, the first director of science in the OECD. King teamed up with the former Fiat motor industrialist Aurelio Peccei and cofounded an influential club of world leaders they called the Club of Rome. One of their first acts was to commission scientists to create computerized forecasts of environmental collapse, which became The Limits to Growth in 1972.
The other was Christopher Freeman, the charismatic University of Sussex economist who was so upset by this idea that he published a stinging rejoinder, Models of Doom.
In 1962 Alexander King and Chris Freeman worked closely together on a very different project, which (at least for King) was a polar opposite to his later work on The Limits to Growth. In contrast to his later incarnation as a skeptic, King was a GDP enthusiast during his OECD days and in fact hired Freeman as a consultant to create a similar accounting system for science spending.
King had been watching closely how GDP was shaping up. As a highly experienced civil servant he understood one of GDP’s great strengths: when something is quantified and valued, and then written up in bright neon lights, it is more likely to be recognized and protected. Most countries in the 1960s, even wealthy ones, were not in the habit of producing accurate accounts of how much they spent, say, on physics projects versus chemistry. King, in his role as head of the science section at the OECD, began to encourage them to do so. At the same time, the OECD itself started publishing league tables of which countries spent the most on their science as a fraction of their GDP.
King and his OECD colleagues also understood something else. Scientists then (and even now) frequently justified their taxpayer dollars and pounds by arguing that all of the leading civilizations needed and funded researchers. King understood that such arguments don’t always wash with ministries of finance. He knew that scientists needed to prove to their governments that money spent on science wasn’t simply a way to fund generous retirement pensions for professors; it was benefiting the economy too.4
King figured correctly that if nations wanted to protect their science spending, they had to find ways to get science mentioned and discussed at the top table of economic policy making, and that would entail adopting some of the language of GDP.
And that is where Chris Freeman came in.
At the request of the OECD, Freeman created an accounting manual for science that helps countries work out how much money goes to different research fields and how much of this spending is helping boost economic growth. This has come to be known as the Frascati Manual (named after the small town near Rome where the first meeting was held). At this event, held in June 1963, Freeman is reported to have said, “Most countries have more reliable statistics on their poultry and egg production than on their scientific effort and their output of discoveries and inventions,” and he recommended that countries should be spending at least 3 percent of their GDP on research, because spending on research helps to boost growth.5
We know from at least that time that the United States wanted science spending to be included in the GDP accounts as an investment. And we know that the United States had good reasons for this. Two thirds of America’s research spending comes from industry, and most of this takes place inside the large multinational corporations, which we know well. For companies such as AT&T, Boeing, Ford, General Electric, and IBM, a dollar spent on research was very much an “investment,” because it more than paid for itself in sales of aircraft, cars, and computers and the profits that accrued from these sales. For America’s landmark corporations of the 1960s and 1970s, it made perfect sense to regard what is known as Research and Development as an investment.
But Richard Stone and his successors handling GDP methodology initially resisted these efforts, not only to classify science spending as an investment, but also to split the science spending figures in a more granular way, and they had equally good reasons for doing so. They knew, first, that research spending in the rest of the world did not (does not) fit the US model: spending on research does not automatically lead to companies making money and therefore achieving higher levels of growth. Apart from a small number of developed countries and the larger emerging economies, science spending in much of the rest of the world is on a much smaller scale and tends to be dominated by the state. Back in the 1960s and 1970s, even the larger developing countries, such as Brazil, China, and India, mostly funded their scientific research from public sources. They did not have research-intensive multinational corporations of the kind we see today. Reclassifying Research and Development from a cost to an investment for these countries therefore would leave no tangible benefit to their national income.
Furthermore, Stone and his colleagues also knew there was an additional problem, and that had to do with data quality.6 Most countries do not accurately measure their science spending to the fine-grained level achieved in richer countries. Countries such as the UK, for example, know precisely how much is spent on physics, how much on the humanities, how much on environmental science. They are also careful to separate research spending that is research for its own sake and research spending that might lead to profits. But outside of a handful of OECD countries, science spending is calculated in a far coarser way. It is usually a single lump sum, often thrown in as part of a larger budget for higher education. Moreover, richer countries don’t collect such data on a quarterly basis, which is what GDP requires. You can see the problems. National statisticians such as Richard Stone were worried that insisting that all of research contributes to GDP growth would open the door to reams and reams of unreliable data. And he wasn’t wrong.
Still, the OECD and the more developed nations didn’t give up, and by 1993 they had persuaded the UN to include R&D statistics inside what are called “satellite accounts.” Satellite accounts are a very interesting phenomenon. They’re a kind of voluntary arrangement whereby national statistics offices promise to start gathering data in an area that governments have begun to regard as important—but not important enough to include in the main GDP accounts. Often, statisticians will use satellite accounts to test whether a lesser-known set of indicators could conceivably make it into the main accounts, as is presently happening with various environmental indicators. Sometimes they’re useful for kicking an unwelcome set of indicators into the long grass. But for a determined government, satellite accounts are the perfect launchpad from which to move something into the main GDP index, which is precisely what the United States and its allies (including Britain) did next.
At the same time that science spending moved into the satellite accounts, a number of the OECD’s research-intensive member states continued to work together. They created a group called the Canberra Working Group, and after many meetings, in 2007 it was agreed that science spending would in the future be treated as an investment and not as an expense. Australia was the first nation to amend its GDP, and it was followed shortly by Canada, Israel, Mexico, and the United States.7
Buried in this story are important lessons for those seeking to change GDP so that it can measure the things that matter to us all. Environmental indicators have made it to the satellite accounts, the outer circle of GDP, just as Research and Development indictators had in 1993. As is the case with science spending, environmental indicators have been part of the satellite accounts for some years. However, the next stage, the jump from the periphery into the main GDP index, could be more difficult unless their proponents understand two realities.
The first is that getting GDP to recognize environmental costs and benefits is not primarily an issue of data quality. Yes, it is true that much of the opposition to the idea that environmental services can be valued in dollars rested on concerns about the quality and accuracy of the figures. And yes, the data does need to get more accurate. Ecological economist Robert Costanza’s many critics repeatedly challenged his team to explain how it’s possible to accurately value something, such as an ocean or a forest, for which there isn’t an army of buyers and sellers, and which isn’t traded on the market. Costanza, who famously valued the world’s ecosystem goods and services at $33 trillion, responded by asking why the same concerns aren’t raised for other components of GDP. GDP’s data quality problems are now legion. But how is it, Costanza asked, that GDP has been allowed to reign supreme for eight decades in spite of them.
We already know that GDP still has to find a way to value volunteering and housework; and we know that many countries still lack the tools to properly measure even its existing components. Even when it is first published, GDP is often revised the following quarter when more accurate figures become available. And we can say with some confidence that serious concerns about data quality did not prevent science spending from moving from the periphery of GDP and into the center.8
The recapitalization of science spending as an investment happened in spite of continuing concerns about data quality. Most countries do not account for their research spending by individual topics or disciplines. Most countries to this day still do not have the capability to distinguish spending for research that is intended to push the boundaries of knowledge and spending for research that is intended to make money. The change happened in spite of the fact that there is little or no benefit for most countries’ GDP. Moving science to the investment column in countries with low levels of R&D spending and weak data-collection systems won’t boost their GDP.
Costanza was right, but his observation was in fact a convenient distraction for his critics because data quality isn’t the main issue. We know from the case of science spending being reclassified as an investment that data-quality concerns are not a primary reason not to change GDP. Most countries of the world still do not collect R&D statistics with anything like the accuracy or the granularity of the OECD nations. But that didn’t prevent a change from happening—a change that everyone needs to implement, regardless of the quality of their data sets or the quality of their data-gathering processes.
The second lesson is to recognize that GDP change will need campaigners to change their approach. Right now perhaps a majority in the environmental community are more supportive of the idea of a dashboard of indicators, including environment, health, and well-being. They remain opposed to a single index and cannot see themselves effectively endorsing a principle they are so opposed to. They are of course right to be skeptical, but they could perhaps reflect on the reasons why so much environment policy they have helped to create has had relatively little real impact. One of these reasons remains the dominance of GDP. If they want to move the world onto a greener path, they will need to engage with GDP’s processes just as much as they do for climate change.
The final lesson is that no revision of GDP can afford to ignore the interests of powerful nations, and especially the United States and China. Nearly fifty years ago the British economist Dudley Seers (and others) was spot-on when he said measurement systems tell you as much about the motives of their designers as they do about what is being measured.
Let’s go back to the original question of why GDP became mainstream in the first place. GDP became mainstream because, as Seers reminded us, it was designed with the interests of rich countries in mind. After the end of World War II, the forerunner to the OECD promoted GDP as a system of accounting to reassure richer nations that the assistance they were providing under the Marshall Plan wasn’t being misspent and was contributing to the growth of economies. Seers and Mahbub ul Haq both understood that GDP, at its core, wasn’t really about benefiting developing countries at all, but they still had no choice but to implement it.
That is why any revision to the index won’t pass muster unless the interests of its founder countries are protected. And first and foremost that means that any revision must not result in a downward slide; any revision to growth cannot result in the large economies shrinking. It doesn’t matter so much if the quality of the data is a bit ropy. What matters more is that countries that are permanent members of the UN Security Council will not allow a change to GDP that leads to them slipping down the league table.
Perhaps Mahbub ul Haq’s greatest error when devising the Human Development Index is that he allowed the United States to slide down the rankings. This was an error he quickly realized and corrected, though by then the damage had been done. Although he fought hard to protect the index team from being interfered with by governments, and especially by the US government, Haq’s victory may well have been pyrrhic. The US administration probably never forgave him for the humiliation of a lowly position in the first HDI tables.
Paradoxically, this error keeps being repeated by the many alternative accounting systems that have been developed since. If valuing the environment or quality of life means richer nations drop down the index, they simply won’t allow the change to happen.
Haq’s heirs cannot afford to make the same mistake again.