Abstract
This Selected Issues paper analyzes the background, stages and developments, and estimation of the direct costs of Paraguay's banking crisis. The paper provides the estimates on the size and evolution of the informal sector, and examines the extent to which the national accounts capture informal activity. The study also estimates the potential output and total factor productivity by examining trends in output, investment, and population growth as well as the direction and size of fiscal impulse on its economy. The paper also provides a statistical appendix report of Paraguay.
II. The Informal Sector in Paraguay1
A. Introduction
This note provides some estimates of the size and evolution of the informal sector in Paraguay and examines the extent to which the national accounts capture informal activity, or measures accurately actual economic activity in general.
The most common definition of informality sets it equal to illegality, which includes any activity failing to adhere to official regulation and standards.2 However, such a definition seems not practical in the case of Paraguay where official regulations often are not observed, yet the underlying activity takes place in the formal economy. An obvious example of this is provided by estimates of tax evasion. A study by FAD reports levels of tax evasion (for 1997) that, according to the above definition, would imply that some 47 percent of total output is produced informally (Table 4).3 Yet it is clear that the criterium of failure to adhere to official regulations, such as not observing tax obligations, would likely overestimate informality, simply because taxes, in part, also are evaded on formal activities.
Informality Suggested by Tax Evasion, 1997
(In percent)
Informality Suggested by Tax Evasion, 1997
(In percent)
| Sectors | Manufacturing | Construction | Trade | Transport and Communication | Services and Utilities | Agriculture and Mining | Total |
|---|---|---|---|---|---|---|---|
| GDP share | 13.9 | 5.3 | 20.6 | 4.8 | 27.7 | 27.7 | 100 |
| Incidence of tax evasion | 60.6 | 93.8 | 57.7 | 26.2 | 43.5 | 29.6 | |
| “informal” GDP share | 8.3 | 4.8 | 11.9 | 1.3 | 12.0 | 8.2 | 46.6 |
Informality Suggested by Tax Evasion, 1997
(In percent)
| Sectors | Manufacturing | Construction | Trade | Transport and Communication | Services and Utilities | Agriculture and Mining | Total |
|---|---|---|---|---|---|---|---|
| GDP share | 13.9 | 5.3 | 20.6 | 4.8 | 27.7 | 27.7 | 100 |
| Incidence of tax evasion | 60.6 | 93.8 | 57.7 | 26.2 | 43.5 | 29.6 | |
| “informal” GDP share | 8.3 | 4.8 | 11.9 | 1.3 | 12.0 | 8.2 | 46.6 |
Thus, to obtain a more accurate gauge of informality, a more narrowly-focused estimation method will be used. In particular, informality will be proxied by estimating the level of output produced by unregistered urban small-scale enterprises and the self-employed—sometimes referred to in Paraguay as the “traditional” informal sector—together with value-added obtained in the reexport sector, centered in and around Ciudad del Este on the border with Brazil.4 Given the relative importance of the latter, it will be scrutinized first. Its contribution to total GDP will then be included in the global estimation.
B. The Reexport Sector
Reexports, or tourist trade, refer to the retailing activities realized in several towns on the border with Argentina and Brazil. Goods are imported into Paraguay, say from Brazil, net of Brazilian sales or value-added taxes, and are then resold to traders (“the tourists”) from the neighboring country. The “tourists”, from Brazil in this case, do not need to declare taxes on their imports from abroad, up to a certain limit, and thus may obtain some products at lower cost in Paraguay than buying them domestically. On the Paraguayan side, while the retailing activity itself is not illegal, little or no Paraguayan taxes are paid on these trades and a large share of the activity goes unrecorded.
Nevertheless, the Central Bank of Paraguay (BCP) has devised some methods to incorporate border trading into the balance of payments and the national accounts statistics. A first method is based on a comparison of official trade data and direction of trade statistics as reported by Paraguayan trade partners. From this, the BCP calculates the value of reexports, assuming a certain amount of domestic consumption and 15 percent value added5 as its GDP contribution. This yields a peak value of 5 percent of GDP for 1995, declining subsequently to 3.3 percent of GDP by 1998.
A second approach developed in a Central Bank study6 arrives at the reexport sector’s contribution to GDP using micro-data. First, the total value of reexports from Ciudad del Este is determined by multiplying the number of tourists and foreign bulk-buyers by their estimated average value of purchases (the latter figure obtained through spot-checking and surveys). Subsequently, assuming a value added on these purchases of 15 percent, the contribution to GDP by Ciudad del Este alone is derived. Finally, taking into consideration that Ciudad del Este makes up about 73 percent of all border trade, the total contribution of reexports to GDP is calculated. As can be seen in the figure below, this leads to a peak value of 10.1 percent of GDP for 1994, dropping to 4.9 percent of GDP by 1998. The midpoint estimate between the two methods, also shown in Figure 2, ranges between 7.4 percent of GDP for 1994 and 1995 and 4.5 percent of GDP for 1998.
Estimates of Value Added in Reexport Activity
Citation: IMF Staff Country Reports 2000, 051; 10.5089/9781451832365.002.A002
Source: Central Bank of Paraguay; and staff calculations.Estimates of Value Added in Reexport Activity
Citation: IMF Staff Country Reports 2000, 051; 10.5089/9781451832365.002.A002
Source: Central Bank of Paraguay; and staff calculations.Estimates of Value Added in Reexport Activity
Citation: IMF Staff Country Reports 2000, 051; 10.5089/9781451832365.002.A002
Source: Central Bank of Paraguay; and staff calculations.The initial rise and subsequent decline of the tourist trade appears to be caused by a number of external factors: to start, Brazil first increased the legal maximum of undeclared imports in 1992, and lowered it again in 1995 and 1996. Also, demand had been fuelled by a strong currency and an economic boom in Brazil in the middle of the decade. Subsequently, regional integration and the harmonization of external tariffs within the Mercosur framework gradually eroded the relative cost advantage of Paraguayan reexports, a process which is likely to continue as the common external tariff (CET) is being implemented incrementally. Finally, and more recently, the economic downturn in Brazil and Argentina since mid-1998, and the loss of competitiveness vis-à-vis Brazil after the devaluation of the real, have intensified this trend.
C. The Size of the Informal Sector: Data from Household Surveys
Tourist trade is, however, only one part of informal activity in Paraguay. Employment and income data from recent household surveys allow an approximation of the total size of the informal sector, including the reexport sector’s contribution derived above.7 Since the household surveys contain data that are disaggregated by type of activity, they allow an alternative estimation of GDP from the production-side. This alternative measure of GDP can then be compared to the official national accounts data to estimate any measurement error in the official data, potentially caused by informal activities.
In the household surveys, the Paraguayan Statistical Office estimates that 46.4 percent of persons in the urban workforce was employed in the informal sector (Li)8 in 1996, and 45.6 percent in 1997–98.9 This information, together with data on the total labor and capital income (y) received by these “informal” workers, permit an estimate of national income generated in the informal economy:
Data on employment in the formal economy, when combined with information on average monthly wages paid in the formal sector,10 also allow for an alternative estimate of national income (as compared with official national accounts data) in the formal economy. To arrive at this alternative estimate, income can be inferred by using a Cobb-Douglas production function:
Assuming that labor is paid its marginal product, formal output can be calculated as:
where (1-α) is the labor share in income. Since data are available on wages and employment in the urban areas by main sectors of economic activity, and since the labor share in income is not uniform across sectors, we estimated formal urban income as the sum of the sectoral results. Moreover, to capture income earned in the formal non-urban, or rural, sectors, the primary sector (agriculture, mining and fishing), as well as secondary and tertiary output produced in rural areas, were also added.11 Tables 2 and 3 present the results for 1996 and 1997–98, respectively.
For 1996 the chosen production function yielded a nominal GDP figure of 124 percent of official GDP, and for 1997–98 of 131 percent. The informal sector contributed 15.6 percent (1996) and 14.6 percent (1997–98) to total output (formal plus informal), or 19 percent of official GDP in both years. These estimates of informal activity are somewhat smaller than those suggested in other studies, and they do not vary significantly over time.12 Moreover, since household survey data are likely to contain some under-reporting of income and activity, the measurement error of official GDP may extend beyond the gap estimated here.
Nominal GDP 1996
(In billions of Guaranies)
Formal plus informal GDP.
Nominal GDP 1996
(In billions of Guaranies)
| Manufacturing | Construction | Trade | Transport | Services | Total | ||
|---|---|---|---|---|---|---|---|
| Differentiated Production Function | a=1/3 | a=1/5 | a=1/10 | a=1/5 | a=1/10 | ||
| Contribution to total GDP 1/ | 2,342.1 | 1,033.5 | 3,891.4 | 1,236.9 | 10,844.1 | 24,463.5 | |
| Contribution to informal GDP | 409.7 | 405.5 | 2,216.1 | 288.7 | 49.6 | 3,819.6 | |
| Informal share in total GDP (percent) | 17.5 | 39.2 | 56.9 | 23.3 | 4.6 | 15.6 | |
| Total GDP as % of official GDP | 123.5 | ||||||
| Informal sector as % of official GDP | 19.3 | ||||||
Formal plus informal GDP.
Nominal GDP 1996
(In billions of Guaranies)
| Manufacturing | Construction | Trade | Transport | Services | Total | ||
|---|---|---|---|---|---|---|---|
| Differentiated Production Function | a=1/3 | a=1/5 | a=1/10 | a=1/5 | a=1/10 | ||
| Contribution to total GDP 1/ | 2,342.1 | 1,033.5 | 3,891.4 | 1,236.9 | 10,844.1 | 24,463.5 | |
| Contribution to informal GDP | 409.7 | 405.5 | 2,216.1 | 288.7 | 49.6 | 3,819.6 | |
| Informal share in total GDP (percent) | 17.5 | 39.2 | 56.9 | 23.3 | 4.6 | 15.6 | |
| Total GDP as % of official GDP | 123.5 | ||||||
| Informal sector as % of official GDP | 19.3 | ||||||
Formal plus informal GDP.
Nominal GDP 1997–98
(In billions of Guaranies)
Formal plus informal GDP.
Nominal GDP 1997–98
(In billions of Guaranies)
| Manufacturing | Construction | Trade | Transport | Services | Total | ||
|---|---|---|---|---|---|---|---|
| Differentiated Production Function | a-1/3 | a-1/5 | a-1/10 | a-1/5 | a-1/10 | ||
| Contribution to total GDP 1/ | 2,774.2 | 1,065.8 | 4,838.7 | 1,813.7 | 14,691.5 | 30,613.5 | |
| Contribution to informal GDP | 628.9 | 350.0 | 2,481.8 | 359.0 | 649.7 | 4,469.4 | |
| Informal share in total GDP (percent) | 22.7 | 32.8 | 51.3 | 19.8 | 4.4 | 14.6 | |
| Total GDP as % of official GDP | 130.6 | ||||||
| Informal sector as % of official GDP | 19.1 | ||||||
Formal plus informal GDP.
Nominal GDP 1997–98
(In billions of Guaranies)
| Manufacturing | Construction | Trade | Transport | Services | Total | ||
|---|---|---|---|---|---|---|---|
| Differentiated Production Function | a-1/3 | a-1/5 | a-1/10 | a-1/5 | a-1/10 | ||
| Contribution to total GDP 1/ | 2,774.2 | 1,065.8 | 4,838.7 | 1,813.7 | 14,691.5 | 30,613.5 | |
| Contribution to informal GDP | 628.9 | 350.0 | 2,481.8 | 359.0 | 649.7 | 4,469.4 | |
| Informal share in total GDP (percent) | 22.7 | 32.8 | 51.3 | 19.8 | 4.4 | 14.6 | |
| Total GDP as % of official GDP | 130.6 | ||||||
| Informal sector as % of official GDP | 19.1 | ||||||
Formal plus informal GDP.
D. A Third Alternative Estimate of GDP—Based on Consumption
The 1997–98 household survey also allows an alternative calculation of GDP from the expenditure side, based on data on families’ average income and consumption and information on the other main expenditure categories. Contrasting the household survey data on consumption with the official figures can provide some further insights as to whether and by how much official GDP is underestimating actual GDP. As can be seen in Table 7, multiplying average yearly consumption expenditure, as obtained in the household survey, by the number of families13 yields a higher consumption figure than what is implied in the official GDP calculation.
Nominal GDP 1997–98, Calculated by Expenditures
(In billions of current Guaranies)
The two estimates vary slightly based on different population data in the national accounts and the most recent household surveys.
Nominal GDP 1997–98, Calculated by Expenditures
(In billions of current Guaranies)
| Private Consumption Expenditure | Total GDP | Adjustment: % of official GDP | Adjustment: % of actual GDP | |
|---|---|---|---|---|
| National Accounts | 15,250.5 | 23,436.9 | ||
| Estimate I 1/ | 19,002.7 | 27,189.1 | 16 | 14 |
| Estimate II 1/ | 19,682.3 | 27,868.7 | 19 | 16 |
The two estimates vary slightly based on different population data in the national accounts and the most recent household surveys.
Nominal GDP 1997–98, Calculated by Expenditures
(In billions of current Guaranies)
| Private Consumption Expenditure | Total GDP | Adjustment: % of official GDP | Adjustment: % of actual GDP | |
|---|---|---|---|---|
| National Accounts | 15,250.5 | 23,436.9 | ||
| Estimate I 1/ | 19,002.7 | 27,189.1 | 16 | 14 |
| Estimate II 1/ | 19,682.3 | 27,868.7 | 19 | 16 |
The two estimates vary slightly based on different population data in the national accounts and the most recent household surveys.
To this, we add government consumption from fiscal data, net exports as calculated by the BCP (which includes informal reexports) and investment. Thus, all expenditure categories of GDP with exception of investment are calculated from independent sources.14 This leads to the result that actual GDP may be up to 119 percent of official GDP, a finding that is close to that found in Part C. Moreover, the increment to official GDP as obtained by the consumption-based GDP calculation is nearly equal to the informal sector GDP contribution.
While the two alternative GDP estimates presented above provide very similar results overall, it would still appear difficult to disentangle how much of official activity, or alternatively, how little of informal activity, is captured in the official income accounts.
Indeed, it cannot be concluded that “formal” activity appears to be captured in full, whereas informal activity is left out completely in the official accounts. For instance, reexports, or at least the lower-bound estimate derived from official statistics, are included by the Central Bank in the official national accounts calculation. Also, the category “other services” in the official national accounts may be capturing part of the “traditional” informal sector. At the same time, illegal activities such as contraband and drug-trafficking are not included in official figures by the BCP, nor are they likely to be declared in the household surveys on income, but they could be influencing the consumption-based calculation. Finally, enterprises included in the formal sector may be under-declaring their output in order to evade taxes.15 Therefore, while the above calculations suggest that that official GDP may be underestimated by between 15-24 percent, this estimate probably reflects a combination of underreporting and informal activity.
E. Conclusion
A conservative estimate of the size of the informal sector, derived from information contained in household surveys, indicates that it amounts to around 19 percent of official GDP (or about 15 percent of total GDP). However, given the basis of this estimate, namely a narrow definition of informal employment, and the fact that people might be under-declaring their income in the household surveys, this might indeed be a lower bound.
The relative size of the informal sector does not seem to have varied much in recent years. Future developments will depend on employment opportunities in the formal sector, labor market deregulation and tax policy. Within the overall informal sector, the reexport sector experienced a boom through the mid-1990s, but has been declining gradually since then, albeit not yet to below historical levels. This development mainly reflects economic conditions in Brazil and Argentina, and Brazilian import policies, as well as the gradual implementation of a common tariff policy in the Mercosur.
With respect to the national accounts, the fact that the alternative calculations of GDP yielded figures for 1996 and 1997–98 between 119 and 131 percent of official GDP could be taken as an indication that total output may be underestimated by this amount. Given our estimates of the informal sector, however, and taking into consideration the BCP’s calculation methodology, it is noteworthy that not only part of informal sector activity may be escaping official documentation. Other omissions might be attributable to contraband and underreporting of formal activity. In any event, the general result that almost half of the urban workforce, engaged in informal activity, is producing less than one sixth of total output seems to corroborate concerns about inefficiencies which need to be addressed.
Prepared by Eva Jenkner.
See Alejandro Portes, “The Informal Economy - Perspectives from Latin America”, in S. Pozo, ed., “Exploring the Underground Economy”, W.E. Upjohn Institute, 1996.
“Paraguay: Estrategia de la Reforma del Sistema Tributario”, IMF FAD, March 1999.
A sizable part of the Paraguayan informal sector may also be made up of illegal activities such as contraband or smuggling. Since due to their very nature little reliable information is available, they are left out of the calculations.
Comprehensive surveys arrive at this level as the average income generated for Paraguayan nationals.
Movimiento Comercial y Financiero de Ciudad de Este, BCP, Noviembre 1998.
Other approaches to estimate the size of the informal sector in an economy include money-demand measures, or indicators of physical activity. Schneider and Enste (1998) for example estimate the informal sector at 27 percent of GDP for Paraguay in 1989-90. (Schneider and Enste, “Increasing shadow economies all over the world - fiction or reality?”, Linz, 1998).
Studies generally define the urban informal sector as consisting of workers in and owners of enterprises with less than 6 employees, as well as self-employed or unpaid family workers.
Trabajadores en el Sector Informal Urbano, Encuesta de Hogares 1996, DGEEC, Paraguay, 1997, and Trabajadores en el Sector Informal Urbano, Encuesta de Hogares 1997–98, DGEEC, Paraguay, 1999, first draft. The latter survey was conducted between July 1997 and July 1998. To simplify the analysis, its results are here compared with data for 1998, which may lead to slight underestimations.
For the formal sector only information on average monthly salaries is available. This information is published in the BCP’s Informe Eeonomico.
Around 30 percent of secondary and 20 percent of tertiary activities are rural and hence make up an important part of formal economic activity.
This partially contradicts anecdotal evidence, which suggests that informality has increased in recent years with rising rates of unemployment (The sum of formal and hidden unemployment has grown from 7.8 percent in 1995 to 14.9 percent in 1998) and continuing rigidities in the labor market. However, one needs to bear in mind that our results are based on employment data which indicated a decrease in the informal sector share in employment in the first place.
Average household size was 4.75 in the survey.
Investment figures are taken from the national accounts. However, they seem reasonably firm and easy to reproduce; Paraguay imports almost all of its capital equipment, and construction volumes are captured reasonably accurately.
A recent study found that a comparison of export data and national accounts statistics imply that 253.3 percent of all clothing or 418.9 percent of all vegetable oil produced are being exported. (Competitividad de las Actividades Industriales, Ramiro Rodriguez Alcala, mimeo).
