I recently came across a marvelous book, Neil Monnery’s “Architect of Prosperity: Sir John Cowperthwaite and the Making of Hong Kong,” a biography of Hong Kong’s financial secretary, who introduced most of the city’s free-market reforms in the 1960s.
“We suffer a great deal today from the bogus certainties and precisions of the pseudo-sciences, which include all the social sciences, including economics,” was Cowperthwaite’s blunt appraisal of his era – and which one would happily apply to our present era, too. “I myself tend to mistrust the judgment of anyone not involved in the actual process of risk-taking.”
One of Cowperthwaite’s bright ideas was to prevent the collection of most economic data for fear that someone might try to do something with it. Any meaningless fluctuation, he reasoned, would encourage bureaucrats to needlessly meddle, while any period of growth might convince businesses that they no longer needed to innovate. When he was once asked how developing countries should improve their economies, he replied: “They should abolish the office of national statistics.”
Should Cowperthwaite’s example be attempted again? Suppose a Southeast Asian government were to ban the collection of most economic data for a few years. Come January, we would not know how much GDP had grown the previous year, nor whether overall trade was up and down, nor which sectors were doing well or poorly. We would have no way of knowing whether its economy was doing any better than its neighbors, nor would foreign investors possess a simplistic (and often faulty) reasoning for picking a certain country. (Perhaps some data needs to be collected for the credit rating agencies and public debt, although such information would ideally be kept from most ministries.)
I think this would have some benefits. Most obviously, it would require far more inventive means of information-seeking. Deprived of spreadsheets, bureaucrats and politicians would be compelled to converse more regularly with workers, managers, investors, and company owners to ascertain what was happening in a particular sector of the economy. From these, they might gather information that aggregated data cannot provide. And in doing so, they might also discover innovative solutions to problems that they never knew existed.
Anecdotal information often trumps statistics when it comes to identifying fixes for future events. After all, most economic data is collected to justify, not to inform; it exists to prove something has worked, rather than explaining why it worked or whether it will continue to work. That a country’s GDP grew by 4 or 5 or 6 percent last year indicates that something went well, but it doesn’t necessarily mean that it will continue to go well or that it won’t produce an unintended consequence. (Running a massive trade surplus made economic sense until Donald Trump came along, for instance.)
Moreover, ever-greater streams of statistics tend to lead to risk aversion, as Cowperthwaite reasoned decades ago. Like computers, most bureaucrats are incapable of thinking – as the British economist John Kay would say – obliquely. Just as SatNavs will only inform you of car routes based on the single metric of speed (rather than simplicity, quality of petrol stations, or bucolicness), once armed with data, bureaucracies focus on specific metrics. This isn’t their fault. Data provides protection. If the existing information suggests a policy or intervention should work and it doesn’t, then you’re going to face much less censure than had you attempted something that went against the data. After all, fail conventionally and you get sympathy; fail unconventionally and you get blamed.
Most economic data also produces odd incentives. Due to our fixation on aggregation, governments are motivated to boost GDP (which is easily measurable) but not individual wealth (which is far more subjective). Take this example. Between 1991 and 2024, Cambodia’s GDP per capita surged from $267 to around $2,500. Although the figures are opaque, I believe I’m right in saying that private debt has increased from almost nothing a few decades ago to at least $3,000 per capita. No economist would even bother asking the question, but it seems somewhat helpful to ponder whether this means the average Cambodian has actually gotten any wealthier over the past four decades of economic growth.
With enough time, one could make a compelling argument that nearly all the financial gains from the last three decades of development have been eroded by the metastasizing extractive, rentier sectors of the economy. However, I don’t think data can ever provide an answer to my question. But anecdotally, you could easily discover whether most people think they have become richer.
H.L Mencken once joked that a wealthy man is one who earns $100 a year more than his wife’s sister’s husband. Perhaps the gag is now a little outdated (although I’m not sure similar competition exists between women and their brothers-in-law’s wives), yet it does highlight that wealth is not only relative but also dependent on perception. I’ve long harbored the suspicion that most Southeast Asian governments would prefer to see Singapore suffer an economic crisis and a whopping fall in GDP rather than for every country in the region to have an okay year.
Likewise, on paper, it’s better to earn $100,000 a year even if you live in an area where everyone else is unemployed than to earn $30,000 but live in a place where your neighbors take home about the same. But the higher-earner walks out of his door to find boarded-up shops and menacing neighbors, whereas the lower-earner can at least expect his neighborhood to have some decent restaurants and cafes. On a good month, I might slip into the global top 2 or 3 percent of earners. But within the country where I live, I’m stolidly middling. Indeed, try telling the average American that they’re part of “The Global One Percent.” It doesn’t compute, since it’s an evolutionary impulse to compare ourselves to our immediate neighbors, not strangers halfway across the world. I cannot find any research on it, but I would be interested in knowing how most people compare their social and financial progress over time. Do people judge their existing level of wealth by how much they had five years ago, ten years ago, or their financial situation in childhood?
Anyway, while some folks argue that instead of GDP, countries should adopt a Bhutan-style Gross National Happiness score, perhaps it’s wiser to stop trying to quantify and metricize and focus instead on anecdotes.