Should America’s GDP data include drug dealing?
Consumer spending on many illegal substances grew rapidly after 1965
Is the US drowning in drug abuse? If you asked most ordinary American voters that question, the answer would probably be “yes”. After all, the past few years have delivered a stream of headlines about the tragic effects of the opioid epidemic, with about 130 Americans dying every day from opioid-related overdoses in 2018 and 2019.
President Donald Trump often rails against the scourge of substance abuse but tends to blame it on imports from places such as Mexico or China. Last month, though, America’s Bureau of Economic Analysis — the branch of government that assembles official data — jumped into the fray. And its conclusions about drug trends might take some observers by surprise.
First, a little background. Until fairly recently, government statisticians assumed their job was to record observable legal economic activity (such as factory output). However, the IMF has long urged them to widen the lens. Since 2014, EU member states have been encouraged by Eurostat, the official statistics body, to include some illegal activities — from prostitution to drug trafficking and tobacco smuggling — in their GDP.
So far, American statisticians have taken a more puritanical approach. “The Bureau of Economic Analysis does not currently include illegal market activity because of challenges in source data and different conceptual traditions,” Rachel Soloveichik, a BEA economist, explained in a paper she presented at an IMF conference last November.
The obvious problem with tracking illegal activity is that it is hard to count with confidence. Soloveichik estimated what might happen if the BEA did include illegal activity in its GDP figures, by amassing all the available sources of data about illegal drugs, prostitution, gambling and corporate theft in the US.
On a macro level, the implications of this experimental exercise are not earth-shattering: if illegal activities were included, it seems total GDP would be about 1 per cent bigger. Judging from Eurostat figures, this suggests that the illegal sector is slightly larger in the US than in some European countries, but not by that much.
But it is the category breakdown that is really intriguing. The biggest illegal category is drugs, which Soloveichik estimates sparked a massive $111bn of expenditure in 2017. However, spending levels have shifted sharply over time. In the 1950s, expenditure on drugs was negligible when compared with overall retail spending. But it then rose sharply: in the 1970s, heroin purchases were apparently equivalent to 3 per cent of all consumer spending, while for marijuana the ratio was 1 per cent. In the 1980s, the rising popularity of cocaine pushed spending on the illegal stimulant to an eye-popping 3.5 per cent.
But after that, spending on most illegal drugs, as a proportion of all consumer expenditure, tumbled so sharply that it now runs at below 0.5 per cent. The only exception to the decline is marijuana: this fell in the 1990s but has since risen slightly, to just over 0.5 per cent.
One explanation for these trends might be a decline in consumption itself, and there are reports that for some drugs, such as cocaine, this is a factor. Another might be that (more recently) legally prescribed opioid drugs have sometimes been substituted for illegal heroin, while marijuana has been partly legalised. But the main cause, according to Soloveichik, is “huge drops in relative drug prices between 1980 and 1990”; indeed the price collapses have been so dramatic that they “may appear implausible at first glance”, she notes.
Nevertheless, this deflation does seem genuine, and it reflects both higher drug potency, increased supply and ease of access. To put it bluntly (no pun intended), over recent decades Americans have been finding it easier and cheaper than ever to get high, even before the recent partial legalisation of marijuana.
This echoes the pattern seen previously with alcohol: during the Prohibition years of the 1920s, expenditure on illegal alcohol peaked at a level equivalent to 4 per cent of all consumer spending, because it was so expensive; after Prohibition ended, total spending levels collapsed because of deflation, even as consumption remained steady.
This has at least two implications. First, it suggests that America’s long-running war on drugs has not “worked” in the sense of making it harder for consumers to access drugs. That does not necessarily prove that decriminalisation would be better, given the damage drugs can do (at last week’s meeting of the American Economic Association in San Diego, for example, a hot topic was the economic damage of the opioid epidemic). But the price trends certainly merit policy debate.
Second, the data shows why we should widen our view when we measure “the economy”. In recent years (as I have previously written), there has been growing concern about the failure of GDP data to properly track digital innovation. But old-fashioned illegal activities clearly matter too; if we exclude either of these, we miss part of the picture. In that sense, the next set of GDP data releases should come with a public health warning; or perhaps a link to Soloveichik’s marvellous paper.