The underground economy: an introduction.

Carson, Carol S.

Recent discussion of the underground economy has raised important questions: How large is it? Is it growing faster than the rest of the econommy? How much tax revenue is lost because of it? Are government policies miscued because it distorts major economic statistics?

The press has explored the underground economy extensively--in the United States, particularly just before the deadline for filing Federal individual income tax returns. Professional groups--including the Federal Statistics Users' Conference, the American Bar Association's Section on Taxation, and the American Institute of Certified Public Accountants--have held conferences or prepared reports relating to the underground ecoonomy. Business groups have focused on it as a consideration in forecasting and planning. Interest has not been confined to the United States. International organizations, particularly the Organization for Economic Cooperation and Development, have sought to clarify issues relating to it. Several international conferences of academic and government researchers have been devoted to it, as was a 1983 conference session of the International Association for Research in Income and Wealth.

This article draws on the variety of materials that is the product of this widespread interest. It has four parts:

* A sorting out of the activities covered by the term "underground economy," a review of incentives to engage in these activities, and a discussion of some definitional issues;

* A synopsis of the various methods that have been used to measure the underground economy or parts of it, and a roundup of results for the United States;

* A survey of the implications of the underground economy, empphasizing its implications for major economic statistics;

* A discussion of the treatment of the underground economy in the U.S. national income and product accounts.

The order of discussion was dictated by several considerations. First, as will become obvious, it is necessary to clarify what is meant by the term "underground economy." The synopsis of measurement methods is useful as background for the roundup of the estimates of the size and growth of the U.S. underground economy. This roundup, in turn, is useful as background for the discussion of implications. The emphasis on implications of the underground economy for major economic statistics leads to the discussion of its treatment in the national income and product accounts.

The first three parts of this article appear in this issue of the SURVEY OF CURRENT BUSINESS. The fourth part will appear in a later issue in order to allow presentation in June of an article that presents material necessary for the evaluation included in the fourth part. The June article will describe the improved adjustments introduced by BEA for 1977 to correct for the misreporting in tax return information used to estimate the national income and product accounts. These two articles have a common purpose: to provide users of BEA's accounts with information needed to judge the strengths and weaknesses of the accounts as they are affected by the underground economy. Further, an article to appear in the fall will present a new indirect method of measuring the growth of the underground economy.

Full citations for the sources mentioned in the text and in the tables of the article in this issue are provided in the bibliography.

I. Underground Activities

The economic activities variously discussed under such catchy titles as "underground," "unobserved," and "hidden" economy are numerous. Some of them are:

* working "off the books" or "moonlighting" (second-jobbing) for cash so that the wages are not reported to tax and social security authorities,

* smuggling,

* illegal gambling,

* working without a necessary permit, as in the case of illegal aliens,

* illegal trade in drugs, tobacco, and alcohol,

* bartering of goods and services,

* do-it-yourself repair,

* padding expense accounts and using office equipment for private purposes (concealing income-in-kind),

* illegal prostitution,

* working while collecting disability or unemployment insurance benefits,

* growing own fruits, vegetables, and other foods,

* loan sharking,

* selling homegrown produce, or homemade items, or personal services that provide income that is not, or is only partially, reported to tax authorities,

* "skimming," that is, pocketing some part of cash-register receipts,

* dealing in land and other assets that yields income not reported to tax authorities,

* working for tips that are not, or are only partially, reported to authorities,

* theft, including theft from business by shoplifters and employees,

* covert rentals.

It is immediately apparent that these activities are quite diverse. Their diversity has several dimensions. First, they include the activities of wage earners, proprietors, investors in real and financial assets, and households--a wide range of economic transactors. Second, most involve production of some good or service. Others, such as theft from households, involve only redistributions of income or property from one person to another.

Third, some take place in the market economy, others outside it. This point is brought out in table 1, which shows, for a selection of the activities just listed, incomes classified as income from market production, on the one hand, from nonmarket production, on the other. Such broad coverage is in line with that outlined by a number of authors--both those developing a conceptual framework and those with a more empirical interest. (See, for example, Feige 1980 and Gershuny, who develop frameworks, and Skolka.) Also, a 1983 international conference on the "shadow" economy included papers that ranged from household production to smuggling. However, work--largely by private researchesrs--has been underway on the measurement and analysis of the nonmarket economy for a number of years (see especially Eisner and Kendrick). Accordingly, the new interest has focused on the market underground. Thus, for this article, the coverage of what will be called the underground economy will be limited to what are generally thought of as market activities; most nonmarket activities are set aside. Of the activities listed earlier as somethimes classified as underground, two will be set aside for this article: do-it-yourself repair and growing own food; barter is a borderline case.

The fourth, and final, dimension of the diversity is also apparent in table 1 for the now narrowed list of activities--that is, those that will be referred as part of the underground. These include activities that are illgal in themselves, as illustrated by income from trade in drugs, and activities that are legal except that the activities or income from them are not reported, as required, to tax, immigration, licensing, or other authorities. This contrast is the basis for the terminology that will be used in this article in referring to income from the underground economy. Incomes from illegal activities are referred to as "illegal-source" incomes. Incomes from activities that are legal--even though tainted with illegality, because they are misreported--are referred to as "legal-source" incomes. These are illustrated in table 1 by off-the-books wages, unreported capital gains, and several others. A few profiles will further identify the kind of situations in which legal-source underground incomes occur.

* A waitress works part-time in a restaurant catering to the breakfast trade; on weekends she helps in an uncle's dry cleaning business. Neither employer withhold income taxes or unemployment insurance. She reports neither the wages they pay, nor the tips she earns, to the tax authorities; she does not file a Federal income tax return. By working completely off the books, she saves both herself and her employers the payment of employment and income taxes.

* A carpenter works weekends for cash, doing home repairs and building porches and other small additions. He counsels prospective employers not to bother with building permits. The carpenter does not report his weekend earnings to tax authorities. He saves the payment of taxes; his employers save the trouble and fee of the building permit and avoid alerting property-tax assessors to the enhanced value of their property.

* An unempoyed writer does freelance editing, but reports that she is unemployed in order to collect unemployment benefits.

* A dentist and a housepainter trade services: The dentist provides braces for the painter's teenage daughters and the painter provides a new interior paint job for the dentist's house. Neither reports as income the value of the services received, and thus both save taxes. They came to this arrangement at a taxpayers' protest meeting, where they agreed that the best way to bring big government under control was to withdraw tax support.

* A semi-retired couple places a "Rooms" sign outside their home in a beach community on summer weekends. They do not report the rents collected to tax authorities.

* The two partners in a prosperous law firm take turns handling the smaller cases--usually wills and divorce proceedings. Neither reports all his income from these cases on his income tax return. Incentives for underground activity

Diverse as underground activities are, they can usually be traced to one or more of several, sometimes interrelated incentives: the desire to evade taxes, the desire to circumvent regulations or prohibitions, the desire to circumvent eligibility and means tests for income support programs; and, perhaps least tangible and separable from the others, the desire to express disaffection with the means and goals of government.

These incentives, which will be explored next, are a necessary but not sufficient condition for participation in the underground economy. The opportunity to participate must also exist, and factors related to opportunity help to explain when and where underground activities occur. Some of these factors are available time (for example, a short enough workweek in the regular economy to permit moonlighting); access to transactions where receipts can escape notice or not leave an "audit trail"; and willingness of the other party in a transaction--for example, one's employee or employer, or buyer or seller--to go along with the evasion or circumvention.

Tax evasion is usually thought to be the most important incentive. In the United States, interest related to the underground economy has centered on Federal individual income taxes. However, other Federal income taxes, unemployment insurance and social security contributions (often referred to as employment taxes), and State and local taxes--such as sales taxes and cigarette excise taxes--are also evaded to various degrees. In other countries, more attention has been paid to evasion of value-added taxes and employment-related taxes (which, in some European countries, amount to about one-half of basic pay). For employment-related taxes benefits of evasion flow to both employer and employee: The employee's wage is free of his share of the tax, and the wage bill of the employer is lower by not paying his share.

In turn, tax evasion can be related to a number of incentives. (For a survey, see Witte and Woodbury.) In general, high rates are an incentive to evade taxes. Further, a person's likelihood of attempting to evade taxes is probably related to his or her perception of the probability of being caught and, if caught, the probability and severity of the penalty. The likelihood of attempting to evade taxes may also be related to the perception of the fairness of the tax system and of whether others--friends and associates, or "everybody"--attempt to evade taxes. In the United States, rising marginal tax rates have increased incentives to hold down tax liability, either by overstating deduction items, understating income, or both.

Further, tax evasion may occur even when it is not the prime motive. A person may, for example, work off the books in order to conceal an illegal immigration stattus. Another factor that contributes to pervasiveness of tax evasion as a feature of the underground economy is that the attempt to evade one kind of tax may necessitate the evasion of others. For example, an employee who tries to evade the income tax may have to evade the social security tax, because employers must report withholding for both income and social security taxes on the same Internal Revenue Service form.

The regulation that motivate underground activites impose a wide range of limitations on the conditions under which income can be earned (or from the employer's point of view, the condition under which workers can be employed). The benefits from circumventing these regulations are added income (or reduced costs) and greater convenience. Two important groups of regulations relate to working conditions and to eligibility for work. In the first group are regulations that stipulate a minimum wage, maximum hours and overtime, safety conditions, and environment protection. In the second are those that stipulate a minimum age or a work permit for aliens. Also, in many jurisdictions, certification is required for the practice of a number of professions--for example, electrician, plumber, doctor and other medical personnel, and beautician. Similarly, licenses are often required for the operation of personal care facilities, such as those that provide child care or nursing. (See especially Tanzi 1983a for mention of additional kinds of regulations in goods, financial, and foreign exchange markets.)

Some potentially income-earning activities are prohibited as inherently contrary to the public interest, and it is the income from these activities that is identified as illegal-source income. Of these, trading in drugs, gambling operations (sports and horse betting, numbers games, casino games, etc.), and prostitution are the most widely prohibited. In the United States, trafficking (that is unauthorized manufacture, distribution, or possession with intent to distribute) in drugs is a Federal offense, and prostitution is illegal in 38 States; in general, these activities, and also gambling, are more widely prohibited in the United States than in other countries. Other widely prohibited activities are loansharking (making loans at exorbitant rates), arson for purposes of fraud, and fencing (trading in stolen goods). In some jurisdictions, the production, distribution, or both, of alcoholic beverages, cigarettes, firearms, and pornographic material are prohibited.

Income support programs may require as a condition for eligibility that a person (or other persons in a household) not have a job, or they may graduate downward a person's benefits as other income rises. An obvious example is unemployment insurance, under which a person must, in some defined sense, be unemployed in order to collect benefits. At present, under the Old-Age, Survivors, and Disability Insurance program, persons who are otherwise eligible for retirement benefits lose $1 of benefits for every $2 they earn over a set amount--$5,160 if they are 65 years old or under, and $6,960 if they are over 65. In cases such as these, the circumvention of eligibility and means tests is a way of securing or maintaining the income support.

Finally, disaffection with the means and goals of government may be expressed in part by refusing to comply with tax codes or reporting requirements of government. The source of the disaffection may be general, for example, the feeling that government has gone too far in replacing individual activity, or that taxes are not being well spent to meet expressed public needs. In other cases, the source may be opposition to a particular policy or project. Increased awareness of the underground economy--even if it is not growing relative to the regular economy--may add to the perception that government is unable to deal effectively with noncompliance of various kinds. Hence, increased awareness may compound the disaffection. Definitional issues

The discussion of activities and incentives to participate in the underground economy point to a broad, general characterization of the underground economy: economic activities--or income from those activities--that elude, wholly partly, a tax or other reporting requirement.

As the focus moves to measurement, such a characterization is not specific enough to be useful. For example, as the Internal Revenue Service has emphasized in explaining the concept of noncompliance, in Income Tax Compliance Research: Estimates for 1973-1981, there is no category labeled "underground economy." "This is because the amounts of unreported income which enter the estimates do not necessarily correspond to any concepts commonly associated with this metaphorical term. . . . The term 'underground economy' . . . has little meaning for tax administration purposes."

Instead, several different definitions are required, specific to the purpose or orientation. This point is brought out in table 1 by the contrast between the illustrative incomes that are in scope for income as defined for Federal individual income tax and those that are in scope for income as defined in the national income and product accounts (for which GNP is a shorthand). The table shows that many of the items in scope for income tax are also in scope for GNP, and vice versa. However, income below the filing threshold, illustrated by the income of an unlicensed beautician who works out of her home, is not in scope for income tax, but is in scope for GNP. Income from illegal activities is in scope for income tax, but not for GNP; the illegality of the activity does not affect liability to pay taxes, but does affect its coverage for GNP because, by definition, GNP excludes them. Orientations other than tax compliance and GNP could be illustrated. For fiscal administration, for example, in-scope underground activities would include, in addition to tax evasion, activities related to circumventing eligibility or means tests in order to claim unemployment or other benefits (see especially van Eck).

The "notes" columns of the table make a further point that income that is not reported on tax returns does not necessarily escape GNP. This point is particularly important because it was often missed in early work on the underground economy; it was mistakenly assumed that, because income tax return information is one of the sources used to estimate GNP, unreported income on income tax returns was unmeasured income in GNP. One reason that unreported income does not escape GNP is that, to the extent that income tax return information is used in preparing the U.S. accounts, adjustments are made by BEA to cope with the misreported income. The table illustrates this reason with the income of the beautician referred to earlier. The second reason is that income tax return information is not the only kind of information used in estimating GNP--nor the incomes associated with it. The table illustrates this reason with covert rent. The estimating method uses information on the size of the dwelling, from a housing survey, and average rent of rented dwellings; neither item depends on a person's reporting his covert income from renting a room in an owner-occupied dwelling. II. Measurement: Methods and Results

Underground activities are not new, but concerted attempts to measure them are. It is already commonplace, however, to say that to measure underground activities is difficult because generally they are meant by those engaged in them to be undetected. Thus, measurement will require, at the least, more ingenuity and, in some cases, different specific data sources than those used to measure other aspects of the economy. Measurement methods

A varity of methods has been used in the United States and other countries to measure the underground economy. The methods can be classified in several ways, including:

* Micro- versus macro-economic,

* Yielding information on compositional detail versus yielding only a single aggregate,

* Yielding information for one or a few points in time versus yielding a time series,

* Direct versus indirect.

The most commonly used clasification is direct versus indirect. As used in this article, direct measures are those that depend on contact with, or observation of, persons possibly involved in an underground activity. Indirect measures resort to some kind of indicator of underground activity. Once the methods are classified in this way, the other possible classifications often are characteristics of the direct and indirect methods, respectively. Direct measures often are micro in their approach and yield information on composition and for points of time; in addition, they are often lower bounds on the size of the underground economy. Indirect measures, in contrast, often are macro in approach and yield a single aggregate and a time series.

Methods of measuring the underground economy are sampled in table 2, where they are classified as direct and indirect. The direct methods shown include both surveys and tax compliance studies. Indirect methods are shown in an number of variant classified according to the kind of inforamtion used: monetary variables, demographic variables, income-consumption relationships, casual factors, and national accounting source data.

As indicated under "coverage" in table 2, the methods sampled include those that measure specific parts of the underground as well as those that measure the underground economy as a whole. One method applies to an illegal activity--heroin distribution; its inclusion highlights the kind of information available for illegal activities. Another applies to the construction industry, which is suspected of being an industry in which underground activity is widespread. Of the methods that measure the entire underground, several arrive at the estimate by measuring first a significant aspect--for example, the hidden labor market--and then evaluating in a more informal way the size of the entire underground. The monetary methods, whose introudction in the late 1970's aroused interest in the underground economy, provide one comprehensive measure. Several variants have been developed; synopses of two are provided.

The methods are drawn from those applied in the United States and in seven other countries. Methods for the United States are over-sampled in order to provide the background for the two following tables, which assemble the measures of size and growth of the U.S. underground. Recently developed or recently extended methods are also over-sampled, because the early methods--particularly the monetary methods--have been widely discussed elsewhere, and also to suggest the directions in which research on the underground economy is going.

The table shows "method" and "results" for each study. Each method has its strengths and limitations. The limitations, as a rule, are severe and--especially for the methods introduced in the late 1970's--have been extensively discussed. For example, the inability of several monetary measures to separate changes in an indicator variable due to the underground from other sources of change is swell recognized, as is the likely downward bias in surveys that depend on self-reporting of underground activities. Strengths and weaknesses arre highlighted in the "comments" (which do not, however, attempt to provide full critiques). (For general disicussion of the various methods, see especially Frey and Pommerehne 1982 and 1984, Havrylyshyn and Woroby, and Henry 1983.) Results: size and growth

Table 3 assembles estimates of the size of the underground economy in the United States in years ranging from 1974 to 1981. The estimates differ in the coverage of the underground they purport to measure and were prepared using a variety of methodologies (all of which were summarized, at least in part, in table 2). The estimates are shown in billions of dollars and as a percent of GNP. (The use of GNP in the percentage calculations is covenient but arbitrary; other aggregates could have served equally well.) Several show separately income earned in legal and in illegal underground activities. As a percent of GNP, legal-source income ranged from 4 to 8 percent. Illegal-source income generally was estimated to be smaller, 1 to 7 percent of GNP. For income earned in the underground economy as a whole, the estimates that are sums of legal- plus illegal-source incomes and another estimate of 8 percent of GNP are clustered in a rather narrow range. Well above them are the estimates of 14-15 percent of GNP and then 20 plus percent ranging up to 33 percent, which are the results of monetary methods.

A range roughly as wide as that for the United States is also apparent in estimates for other countries. The estimates shown in chart 3 also vary in methodology used, coverage of the underground they purport to measure, and time period covered. The 33 percent of GNP that was the high estimate for the United States is equaled only in Italy, for which the estimates range down to 10 percent. Anecdotal evidence suggests that the underground economy in Italy is likely to be at the high end of the range for industrial countries. For Germany, the estimates range from 2 to 12 percent of GNP, and for Sweden, from insignificant to 17 percent. Sweden is of particular interest because taxes and social security contributions take and especially large share of income and because regulation is extensive, but, on the other hand, the social fabric is tightly knit. Outside Europe, the estimates for Canada range from 5 to 22 percent, those for Australia range from 3 to 13 percent, and those for Japan from 4 to 15 percent.

The rate of growth of the underground economy, particularly in comparison to that of the measured economy, is perhaps of even more interest than its absolute size. Table 4 assembles various estimates of the average annual rate of growth of the underground economy over various periods from 1974 to 1981 and, in the addendum, the rate of growth of measured GNP. For both 1974-80 and 1976-80, measured GNP grew at an annual rate of 11 percent. Estimates of the growth of the underground range from slightly less than that of GNP to more than that of GNP--14 percent (two estimates)--and substantially more--19-20 percent. A well-publicized estimate by Feige for 1976 and 1978 implied an annual rate of growth between those years of 38-55 percent. (Later work by Feige included a time series, but only in chart form.) For perspective on the Feige estimates, the growth of the underground over this shorter period was calculated for the two time-series estimates by Gutmann and Tanzi; the rates shown were much lower, 10 percent and 19 percent. Results: an eclectic sketch

Various studies of the underground economy or parts of it--those summarized in table 2 supplemented by others--can be pieced together to provide an eclectic sketch of the dominant features of the part of the U.S. underground economy that yields legal-source income. These activities account for the bulk--at least one-half and up to as much as three-fourths--of the U.S. underground economy, although it is not clear whether they are the faster growing part of it or not. (Of the estimates in table 4 that show legal- and illegal-source income separately, one shows legal-source income growing faster and the other shows it growing slower.)

The opportunity to engage in these activities is greatest when the transaction is visible only to two parties. More specifically, opportunity is greatest when cash (or other goods and services, as in barter), rather than check or credit card, is used in payment and when there is little other evidence of the transaction. The situation arises most commonly in small operations--small especially with regard to the number of people (for example, a proprietor working alone or with one or two assistants), but sometimes also with regard to the amount of time (temporary or seasonal work), capital investment, and transaction value. The industries in which these conditions are widespread inlcude trade, services, construction (especially residential), and farming. (See Feffer et al. for a case study examining the construction industry to see in what kind of transactions the opportunity to engage in underground activity arises.)

A wide range of goods--often consumable goods and specialty products--and services are involved in what the Internal Revenue Service has called the "core" of the underground--a variety of informal, often cash-related, arrangements. These goods and services include home repair and additions (carpentry, painting, etc.), food and catering, child care, lawn maintenance, domestic service, and auto and appliance maintenance (see Smith, Moyer, and Trzcinski, and Ferman and Berndt).

If estimates of noncompliance with Federal income Tax laws are taken as indicative of the underground economy, one-third of the underground's legal-source income is in wages and salaries and another one-third in income of the self-employed, that is, professionals like doctors and lawyers, other nonfarm proprietors, and farmers (table 5). For wages and salaries, where there are both withholding and information returns (W-2 forms), income not reported represents only about 6 percent of the wages and salaries that should have been reported; about 94 percent of wages and salaries were voluntarily reported on tax returns. For self-employment income, where there is no withholding and where limited requirements for information returns were put in place only recently, about 41 percent was voluntarily reported.

Who participates in the underground? If estimates of noncompliance with tax laws are again taken as indicative, participation is rather widespread: Surveys show that 20 to 25 percent of the people interviewed admit to some kind of noncompliance. Persons who are younger, in higher and lower (rather than middle) income groups, are self-employed, and have more education reported lower compliance levels than others (see Write and Woodbury). Other survey evidence indicates that one of five households has at least one member engaged in some informal--that is, "on the side"--way of earning income (see Smith, Moyer, and Trzcinski).

Work in the underground may be either full- or part-time, and may be either a person's only work or may be in addition to work in the regular economy. Of informal suppliers to consumers, about one-quarter had regular jobs, so that their underground income supplemented regular income (see Smith, Moyer, and Trzcinski). Workers include not only those employed in the regular economy, but also persons currently unemployed in the regular economy and persons--such as children and retirees--not officially in the labor force. Work in the underground cuts across racial, ethnic, social, and occupational groups) see Ferman and Berndt).

Underground participants have several kinds of buyer-seller relationships with the regular economy. Underground producers buy and consume the variety of raw materials and services produced in the regular economy. For example, an artist who sells a painting (but does not report the income) buys his canvas, paint, and brushes in the regular economy. Underground participants also distribute (and redistribute) and repair the products of the regular economy. Some of the goods produced, furthermore, are sold in the regular economy; the painting just mentioned might be sold to a gallery that resells it in a wholly regular way (see Ferman and berndt).

The sketch that emerges of the part of the underground engaged in the production of legal goods and services--of the kinds of transactions, of the number and variety of participants, and of the kinds of work and buyer-seller relationships maintained with the regular economy--confirms the diversity first noted with regard to the list of underground activities at the beginning of the article. It suggests that, even for this part of the underground, reference to it as an "economy" should not be taken to imply more unity within itself and separateness from the regular economy my than is actually the case.

The information on which this sketch is based is incomplete in many ways and further research will probably show that is is flawed as well. The information on the part of the U.S. underground that yields illegal-source income is even more limited. Until the last few years, quantification had centered on drug trafficking, prostitution, and gambling (which are mentioned in descending order of size). For surveys of illegal activities, see Simon and Witte, and Abt Associates, Inc.) Trafficking in drugs was best documented, in part reflecting concern from a law enforcement point of view with its rapid growth. Research is now extending into other illegal areas. An important part of research--for the legal part, as well--will be to develop further the methods summarized in table 2 that provide information on the composition of underground activities: who, where, how, and under what circumstances.

III. Implications

The existence of a "sizable" underground economy and one that may be growing relative to the regular economy has several implicatons. The implications for major economic statistics and policy based on them will be explored following brief mention of three other areas.

First, because a major part of the activities in the underground economy involves tax evasion, loss of public revenue is clearly an important implication. The revenue loss associated with the $132 billion (mentioned in table 3) in unreported legal-source income on Federal individual income tax returns in 1976 was $35 billion--about one-quarter of collections from the individual income tax. The revenue loss projected for 1981 was $75 billion. This loss raises issues of equity and efficiency, and losses on other Federal taxes and on States and local taxes have the same effect. For a given level of taxes, the rate on reported income will have to be higher; that is, the burden of providing revenue to support the services provided by government to all--including the underground--is carried by those who to not evade taxes. The higher taxes on reported income further distort the allocation of resources between taxed and untaxed activities.

Second, to the extent that the underground economy may be growing relative to the rest of the economy, there is an implication that laws and government regulations are increasingly being flouted. Especially in a country, such as the United States, that depends to a substantial extent on voluntary compliance with tax laws, this implication may point to the need to reexamine tax laws and enforcement strategies.

Third, and closely related to the second, are implictions that stem from the existence and relative growth of a part of the economy that operates in a way different from the regular economy in some important respects. On the one hand, underground activities tend to rely on less efficient information systems--word of mouth, for example--than the regular economy and may use less up-to-date, or a narrower range of, technologies. On the other hand, the underground may introduce flexibilities--part-time and at-home work, for example--not generally available in the regular economy. In turn, these differences can be viewed in several ways. Viewed as an advantage, the underground provides a social "safety valve" for unemployed or underemployed workers, or, from the point of view of the individual, it may be a place to get the experience to break into a job in the regular economy. Viewed as a disadvantage, growth of the underground may reduce social cohesion, for it represents a part of society that--at least to some extent--chooses to set itself apart (see especially Ferman, Berndt, and Selo; contini; and Hansson). Implications for economic statistics

The underground economy has possible implications for a wide range of amcroeconomic statistics. To the extent that income and production in the underground economy are missed, the Nation's production as measured by GNP and national income would be understated. To the extent that jobs in the underground economy are missed employment and labor force statistics would be understated. Statistics on saving and on productivity are also cited as being understated; those on unemployment and the unemployment rate are cited as being overstated. Price series are said to be affected, but those who claim mismeasurement of the rate of inflation disagree about the direction: Some believe that the rate of inflation is overstated, and others that it is understated. Further, statistics on income distribution and on the international balance of payments are cited as being mismeasured. (See, for example, Gutmann 1983, Simon and Witte, Reuter 1982, and Feige 1979.)

The implication for policy based on these measures is clear: The policies--fiscal, employment, industrial, and international, among others--may be responding to, and may be designed in the light of, statistics that give distorted pictures of the economy. It is alleged by some that the economic situation in 1978-79 may have been such a case. Consumer debt burden was one of the indicators that suggested the imminent onset of a recession; debt was so large relative to income that further expansion of consumer spending seemed unlikely. The recession came, but it came later, and did not last as long as expected. It is alleged that debt burden miscued forecasters. According to this view, counterrecessionary monetary and fiscal policy would have been based on a distorted picture of economic developments. Had measured income included income from underground activities, debt burden would not have flashed a danger signal (see, for example, Molefsky).

Although the underground economy clearly has implications for economic statistics, it is not correct to imply--as has often been done--tht the size of the underground is a useful guide to the extent of the possible under- or over-statement of economic statistics. For example, the likely overstatement of the unempolyment rate is probably not as large as some indicators of the size of the underground economy might suggest. The measured unemployment rate prepared by the Bureau of Labor Statistics and a "true" rate that takes the underground into account could differ; the possible difference depends on whether underground workers respond to questions asked in the survey used to determine the unemployment rate, how they respond (truthfully or not), and how they are employed (part-time or full-time, and only in the underground economy or also in the regular economy). If undergrougn workers respond, but rather than reporting their only and full-time job in the underground, they report that they were unemployed, the measured unemployment rate would be overstated. However, this combination is not the only possible one. If they responded that they were employed, as they might if they found it easier to use work in the regular economy as a cover for work in the underground, the measured unemployment rate would not be overstated; it would be the same as the "true" rate. In addition, information about unemployment as measured by claims for benefits, where there is additional incentive to report as unemployed, is not directly transferable to the measure of unemployment just referred to. Unemployment is defined differently for the two series and tabulated as parts of two separate operations. (See McDonald for an examination of the effect of the U.S. underground economy on the Bureau of Labor Statistic measures of the labor force, the Consumer Price Index, and productivity.)

Further, for many policy-oriented uses of economic statistics, change over time is more critical than level. Use of GNP is an example: A low or high percentage rate of growth is more likely to be a factor in policy determination than is the billions-of-dollars level. Thus, not only the size of the underground economy, but also its rate of growth, are relevant in evaluating the extent to which the underground economy has led to miscuing of policymakers. The estimates of the growth of the underground economy are even more tenuous than are estimates of its size, a fact that suggests that even more caution be used in drawing conclusions about the possibility of miscuing by economic statistics.

Each set of economic statistics--employment and unemployment, prices, national income and product, and so on--must be evaluated separately, because differences in scope and in the sources and methods used to prepare them mean that the underground economy affects them differently. In the next part, the scope of GNP, national income, and other major measures in the U.S. national income and product accounts will be reviewed and the sources and methods used to prepare them evaluated in light of the underground economy. A situation mentioned earlier highlights the usefulness of this approach. In several countries, including the United States, tax return information is one of the sources used to prepare the income estimates in the national accounts. Its bias due to tax evasion has long been recognized; other things equal, it is used only when other source information thought to be less biased is not available. Further, when it is used, adjustments are usually made to take a likely degree of misreporting into account. (These adjustments to the U.S. estimates have been improved for the year 1977, and will be discussed in the June issue of the Survey.)