Though policymakers, researchers, journalists, and others speak frequently of the underground economy, they often talk past each other because underground economic activity has not been clearly defined. Is it the production and distribution of illicit goods and services, such as crack cocaine and prostitution? Does one instead mean economic production that is simply unaccounted for because the producers wish to avoid paying taxes—as in the case of a bartering transaction between a dentist and her manicurist? Or are we referring to home production—the value of output from the family garden plot?

These belong to different categories of economic activity. But what is common among all three is that they are difficult to measure systematically and often elude those drawing up the national income and product accounts. In this respect, official estimates of national income understate the true level of output.

To what extent is it a problem that we do not account for all economic activity? The answer depends on the activity’s character. The implications of not accounting for home asparagus production are fairly inconsequential. In contrast, knowledge that the illicit drug trade has doubled has longstanding economic and social implications. Likewise, tracking growth in the level of tax evasion and understanding the degree of unreported economic activity that takes place are important. Increased tax evasion will result in a declining tax base, which will undermine the fiscal authorities’ ability to finance infrastructure projects. Unregulated production, meanwhile, may have important consequences for worker safety. In this respect, the measurement of illicit economic activity and of unregulated and untaxed economic activity is vital. In what follows, I limit my discussion to these two broad categories of underground activity: illicit economic production and unregulated or untaxed economic production.

Clues and Traces of Underground Economic Activity

To measure the magnitude of underground activity, we can sum estimates of unrecorded economic production for each sector. However, such an analysis requires an inordinate amount of specialized information about illicit drug production, illegal prostitution, the loan sharking industry, tax evasion, and so forth. As such, the sector-by-sector approach is suboptimal if we wish to obtain time-series and comparable cross-country estimates. Alternative techniques are more easily applied to many countries and many time periods.

Economic production of goods and services leaves traces on the economy beyond the product or service itself. The mining of ore leaves tailings, and the production of irrigated crops decreases aquifer levels. For those of us interested in accounting for underground economic activity, these by-products constitute our main source of information. By analyzing their behavior, we derive conclusions about changing magnitudes of illicit, unregulated, and untaxed economic activity. We resort to a variety of techniques to uncover the traces left behind.

First, changing labor force participation rates may indicate underground activity. If individuals are trying to “hide” their participation in illicit or unregulated economic activity, or if they are engaged in tax evasion, they may decline to admit that they are in fact working. Increases in underground economic activity would then be reflected in labor force participation rates. In the United States, the labor force participation rate for men over age 20 is currently reported at about 75 percent. In 1970 the rate was much higher; about 83 out of every 100 males over age 20 worked. The eight-percentage-point decrease in official labor force participation is sometimes attributed to a shift in work from the formal economy to the underground economy. There are concerns, however, with this conclusion. Other plausible explanations may account for the large drop in recorded labor force participation: changes in family structure, increased labor market opportunities for women, and changes in retirement patterns are equally compelling reasons.

Second, currency studies capture the variety of ways in which economic agents effectuate transactions. We pay for goods and services using currency, checking and debit accounts, credit cards, and a variety of other means. In the United States, the level of liquidity (currency circulating outside banks plus checkable deposits) is controlled by the Federal Reserve, but its distribution into currency and checkable deposits is determined by the habits and desires of the public. If individuals wish to hold more of their liquid assets as currency, they will transform checking deposits into currency by making withdrawals from their checking accounts. If individuals wish to hold less currency, they make deposits and will reduce the currency checkable deposit ratio. Liquidity remains the same, but the distribution into currency (outside banks) and deposits varies.

Individuals who transact in the underground economy will presumably prefer to use currency because cash transactions can remain anonymous. An examination of currency in circulation might therefore track increases and decreases in underground activity. According to the Treasury Bulletin in 1960, currency circulation in the United States was US$177 per capita. Today, adjusted for inflation, US$387 circulates for each person in the United States—currency holdings have more than doubled. Does this change reveal an increase in illicit and unrecorded transactions?

Not necessarily. The ratio of currency holdings to other forms of liquidity, not the absolute level of currency, is what matters. In this case, the currency deposit ratio confirms that there has been a substantial increase in currency relative to deposits, from about 0.38 in 1960 to 2.2 in 2005. Still, many causes other than the underground economy can explain changes in this ratio. Perhaps most important, the US dollar has gained popularity over the years as a means of conducting transactions in other countries. When monetary systems “malfunction” in other parts of the world, currency substitution often results in the use of US currency. Much of the increase in demand for US currency derives from individuals abroad seeking a substitute for their own poorly functioning currency. Hence, much of the rise in US currency deposit ratios has nothing to do with US events.

A third approach to measuring underground economic activity uses statistical advances in measuring unobservable or latent variables. This approach presumes that there are many different propagators of underground activity, including burdensome regulations and societal attitudes regarding governance. All of these are, to various degrees, measurable. There are also a multitude of effects of underground activity, some of which we have already discussed—including currency use and labor force participation rates. In the unobservable variables technique, statistical methods relate these “propagator inputs” to the observable “outputs,” deriving estimates of the growth of underground economic activity.