Mexico: Selected Issues
Author: International Monetary Fund
Publication Date:
23 Dec 2004
eISBN:
9781451825626
Language:
English
Keywords:
ISCR; CR; interest rate; PRONAFIDE; deficit; export share; business cycle; Mexico's export; exports to the United States; nonoil export; Export performance; Real exchange rates; Trade balance; Manufacturing; Europe; IMF and Bank
Download PDF (612 KB)

This Selected Issues paper analyzes the sources of Mexico’s economic growth since the 1960s, and compares various decompositions of historical growth into trend and cyclical components. The role of the implied output gaps in the inflation process is assessed. The paper presents medium-term paths for GDP based on alternative productivity growth rates. The paper also describes the significant steps Mexico has taken to strengthen the structure of its public debt in recent years, both in terms of currency composition and maturity.

Abstract

This Selected Issues paper analyzes the sources of Mexico’s economic growth since the 1960s, and compares various decompositions of historical growth into trend and cyclical components. The role of the implied output gaps in the inflation process is assessed. The paper presents medium-term paths for GDP based on alternative productivity growth rates. The paper also describes the significant steps Mexico has taken to strengthen the structure of its public debt in recent years, both in terms of currency composition and maturity.

IV. Mexico’s External Trade: Recent Developments and Policies1

Abstract

The paper examines developments in Mexico’s exports during 1980-2004. There is little evidence that the relationship to U.S. activity has broken down in recent years, or of export displacement by Mexico’s competitors. The recent loss of U.S. market share appears to be related to a strong reaction to the U.S. business cycle and changes in the composition of U.S. demand, and could be partly reversed during the current expansion. Simulations of the impact of various possible trade agreements on Mexico suggest that their effect would be generally small, with the largest impact coming from the phasing out of global Multi Fiber Agreement.

A. Introduction

1. As a result of trade liberalization in the late 1980s and then entry into NAFTA in 1994, Mexico has been transformed from a commodities exporter to an exporter of manufacturing products. Fast growth in manufacturing exports—between 1990 and 2003 the U.S. dollar value of nonoil exports increased almost 4.5 times—slowed markedly in 2001, however, and began to recover only in early 2004 (Figures 1 and 2). During the same period, Mexico grew more dependent on the U.S. market, which is now the destination for more than 85 percent of Mexico’s nonoil exports. This paper discusses recent developments, examining the factors that have affected Mexico’s performance, and assessing the possible impact of future changes in international trade agreements (CAFTA, Multi Fibre Arrangement, etc).

Figure 1.
Figure 1.

Mexico, China, and Canada: Export Developments, 1980-2003

(In percent)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: WEO and OECD databases.
Figure 2.
Figure 2.

Mexico: Nonoil Export Developments, 1994-2004 1/

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: IMF1/ All export data are c.i.f. and seasonally adjusted.Nominal variables; 6-month moving average.

2. The paper finds a limited impact of Mexico’s competitors on its export performance—in particular, there is little evidence that China’s exports have displaced those of Mexico. The recent slow growth of Mexico’s exports is explained in a VAR model by a strong reaction to the U.S. business cycle and changes in the composition of U.S. demand. Simulations using the Global Trade Analysis Project (GTAP) model suggest that the impact of various free trade agreements on Mexico will be small, while structural reforms raising productivity in manufacturing would have a profound impact on exports and growth.

B. Recent Export Performance and Stylized Facts

3. Although Mexico’s post-1980 trade performance has been remarkable, the slowdown since 2000 has highlighted the country’s vulnerability to swings in external demand (Figure 1, upper panel). At the onset of the 1980 debt crisis, Mexico exported less than 10 percent of its GDP, and most of those exports were primary commodities (Figure 1, middle panel). By the end of the 1990s, Mexico exported close to 30 percent of GDP, with more than 80 percent of total exports being manufactured goods (Figure 1, bottom panel). However, export expansion did not translate into a marked acceleration in economic growth,2 while Mexico grew increasingly dependent on the U.S. market (Figure 2, upper panel). The share of Mexico exports to the U.S. increased to about 85 percent in 2002-03, and those exports remained highly correlated with the U.S. business cycle.

4. The annual rate of growth of total and nonoil exports averaged 10 percent and 10 ½ percent during 1980-2000, respectively, and these rates were close to those of China. As a result, during this period, the export shares of Mexico and China grew in tandem (Figure 2, middle panel). Somewhat surprisingly, membership in the North-American Free Trade Area (NAFTA) appears to have no lasting visible impact on Mexico’s rate of nominal export growth (Figure 2, bottom panel).3

5. Much of Mexico’s exports remain concentrated in the maquiladora sector.4 These firms generally have a high import component of their inputs and high labor-to-capital ratios. Maquiladora firms, typically stationed across the border from the United States, import semi-finished products and, after using mostly low-skilled labor, export finished products back to the United States. Maquiladoras tend to be concentrated in the automotive and electrical/electronic sectors (Table 1). The share of gross maquiladora exports in nonoil exports rose from one-third in the early 1980s to more than one-half recently.

Table 1.

Largest Maquiladora Enterprises in Mexico, 2004

article image
Source: Maquila Portal, http://www.maquilaportal.com/cgi-bin/top100/top100.pl.

6. Following China’s entry into the World Trade Organization (WTO) in December 2001, its exports to the United States grew on average 6 times faster than those of Mexico (Figure 2, bottom panel). At a disaggregated level, in 2002 and 2003 Mexico experienced negative export growth in twice as many commodity groups as China (based on the 5-digit end-use classification).54 Moreover, in Mexico the contracting commodity groups were in industries that constitute a significant part of Mexico’s exports to the United States—approximately 25 percent of total exports (electric appliances, computers, telecommunications equipment, cars, and apparel). In contrast, China’s export share losses were in sectors that contributed less than one percent of total exports.

7. Evaluating other key exporters to the U.S. market, Mexico fared better than Canada and Japan, whose exports declined in about 40 percent of commodity groups. More important, contractions in several of Japan’s export commodity groups (such as electric apparatus, machine tools, computer accessories and peripherals, telecommunication instruments, semiconductors, and engines) were matched by similar increases in China’s commodity groups. Until 2001, the combined shares of Japan and China in the United States market were fairly stable at around 14 percent; they then increased to about 18 percent only after China’s entry into the WTO. These results may indicate that, as real appreciation of the yen has priced Japan’s exports out of the U.S. market, Japanese firms have increasingly moved their manufacturing processes into China.

8. Trends in the real exchange rate and growth in nonoil exports suggest a possible long-term effect of relative price on exports (Figure 3, upper panel). Real appreciation of the peso from 1996 until end-2001 was associated with a slowdown in export growth, while real depreciation from 2002 on has been associated with some acceleration of exports. Moreover, domestic nonoil producer prices grew faster than exports prices (Figure 3, bottom panel). In the next section we will explore these relationships in multivariate regressions.

Figure 3.
Figure 3.

Mexico: Exports, the Real Effective Exchange Rates, and Prices, 1994-2004

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: IMF and Bank of Mexico.

9. The external environment is unlikely to get less competitive, and maintaining a steady rate of export growth will be a challenge for Mexico. While some industries will always retain the advantage of proximity to the U.S. markets,55 the future of other industries will depend on macroeconomic and structural polices and infrastructure improvements.

C. An Empirical Model of Mexican Exports

10. In this section we present a VAR model linking the share of Mexican nonoil exports in the U.S. market to the U.S. business cycle, the composition of U.S. demand, a measure of real exchange rate, and China’s and Japan’s exports. The hypotheses to be tested are as follows:

  • Mexico’s exports follow closely the U.S. business cycle;

  • The composition of U.S. demand matters in determining U.S. market shares, as Mexico exports mostly intermediate products, while China and Japan export mostly final consumer products;

  • Appreciation of the real exchange rate slows down export growth;

  • China has displaced Mexico’s exports to the United States

11. The empirical findings suggest that Mexican exports are closely tied to U.S. output, and that their post-NAFTA export expansion was largely unrelated to competitiveness as gauged by a real exchange rate measure. Moreover, we find little support for the hypothesis that Mexico’s exports are being displaced by China’s exports. The slowdown in Mexico exports to the United States during 2001-04 is explained, instead, by the U.S. recession and the relative decline in U.S. demand for intermediate products vis-à-vis consumer products. At the same time, following the entry of China into the World Trade Organization in December 2001, the growth rate of its exports increased without disturbing the Mexico-U.S. export link.

Methodology

12. This study uses a VAR model to characterize trade developments among the U.S., Mexico, China, and Japan. The processes are modeled by

A0xt=Σi=1nAixt−i+ɛt⁢  ,

where xt is a vector of endogenous variables—the trade shares in the U.S. market of Mexico, China, and Japan—and εt is a vector of uncorrelated white noise structural disturbances. In addition, the VAR captures effects exogenous to the system, namely the U.S. output gap, and real effective exchange rates. Finally, we included an intercept dummy capturing the impact of China’s entry into the WTO. The system is estimated using quarterly data from 1980(1) to 2004(1). Regarding lag length, both the Akaike and Schwarz tests indicated three lags.

Data

13. The variables (Mexico’s series are plotted in Figure 4) and their sources are as follows:

Figure 4.
Figure 4.

Mexico: Variables Used in the VAR Model, 1980-2004

(Variables in first differences)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: IMF, U.S. Department of CommerceDeviation from the Hodrick-Prescott filter.A ratio of industrial production and GDP indexes.

  • Exports—seasonally adjusted shares of Mexico, China, and Japan nonoil c.i.f. exports to the U.S. (source: U.S. Department of Commerce);

  • U.S. output gap—index of industrial production, deviation from the Hodrick-Prescott filter. (source: Federal Reserve System);

  • Composition of U.S. demand—the ratio of U.S. industrial production to GDP. (source: Federal Reserve System);

  • Price level—seasonally adjusted producer price indexes. (source: International Monetary Fund);56

  • Nominal exchange rate vis-à-vis the U.S. dollar. (source: Information Notice System, International Monetary Fund);

  • A time dummy to capture the impact of China’s entry into the World Trade Organization (zero until 2001(3) and one afterward).

The model was estimated in first differences, as the variables in question are nonstationary (Sims et al,1990).

14. The VAR analysis of exports to the U.S. presents some special difficulties. First, Mexico completed its gradual move away from protectionist policies with entry into NAFTA. The process of regime switching was slow rather than abrupt, and thus is not consistent with the use of binary dummy variables.57 Second, the Tequila crisis at end-1994 brought about significant economic and structural changes that accelerated Mexico’s integration with the U.S. economy, accelerating export growth temporarily. However, a Tequila crisis dummy variable was statistically insignificant. Third, the rise of China and its integration into the world economy created other shifts in trade patterns (International Monetary Fund, 2004). Driven by steady reduction of trade barriers, high investment, fast labor productivity growth, and the availability of a large labor pool, China exports to the U.S. expanded from ½ percent of total U.S. imports in 1980 to 14 percent in early 2004.

Estimation Results

15. The results for Mexico exports from the VAR estimation are broadly consistent with earlier literature,58 as well as with the relationships observed in Figure 1 and 2. They seem to contradict, however, the relationship between the real exchange rate and export growth. Summary statistics are satisfactory: the model explains about 55 percent of variability in the differenced endogenous variables and an F-test on the regressors fails to reject the model. The actual and fitted export shares for Mexico, China, and Japan are shown in Figure 5 and the latter seem to explain reasonable well the recent development.

Figure 5.
Figure 5.

Mexico, China, and Japan: Actual and Fitted Trade Shares, 1980-2004

(In percent of U.S. imports, first differences)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: Author’s calculations.

16. Impulse response functions (IRF) show how the system reacts to one-standard-error shocks to the key variables: the results for Mexico, China and Japan are shown in Figures 6, 7, and 8, respectively. Overall, the estimates show that the autoregressive part of the system is strong for all countries—an increase in own-country export share remains broadly stable for the duration of the simulation (16 quarters).

Figure 6.
Figure 6.

Impulse Responses of Mexico’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: Author’s calculations.
Figure 7.
Figure 7.

Impulse Responses of China’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: Author’s calculations.
Figure 8.
Figure 8.

Impulse Responses of Japan’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)

Citation: IMF Staff Country Reports 2004, 418; 10.5089/9781451825626.002.A004

Source: Author’s calculations.
Country results

17. Mexico. First, Mexico’s exports to the United States follow closely the path of the U.S. output gap. Second, we find that during industrial production slowdowns in the U.S. economy—during which the IP-to-GDP ratio declines—the share of Mexico’s exports declines, as its exports are more tied to industrial production. Third, the IRF of the real exchange rate is very small, suggesting that the quantitative role of price and exchange rate shocks was limited during the period under consideration. Fourth, we find little evidence that China exports displace Mexico exports—on the contrary, the IRF suggests that a rise in China’s share is associated with an increase in Mexico’s exports. In contrast, the IRF suggests that Mexico may have displaced Japan’s exports to the United States.

18. China. First, China’s share appears inversely related to the U.S. business cycle, as measured by the deviation of U.S. industrial production from a trend, reflecting the steady increase in the China’s exports. Second, similarly to Mexico, China’s exports appear positively related to the U.S. IP-to-GDP ratio. Third, a positive shock to Japan’s export share tends to reduce China’s shares. Fourth, an increase in China’s competitiveness, that is, real depreciation, is associated with a modest increase in the export share. Finally, the time dummy variable for China’s entry into the WTO is statistically significant.

19. Japan. First, Japan’s share appears inversely related to the U.S. business cycle, as measured by the deviation of U.S. industrial production from a trend, reflecting the steady decline in the Japan’s exports to the U.S.. Second, and in contrast to Mexico and China, Japan’s exports—mostly final consumer products—appear negatively related to the U.S. IP-to-GDP ratio. Finally, real appreciation of the yen is associated with long-term losses in the export share.

What explains the recent weakness in Mexico’s export share?

20. In the context of this model, two main factors explain the relative weakness in the Mexico’s export share in 2001-03. First, Mexico’s reaction to the U.S. business cycle is strong and opposite to that of China. Second, the change in composition of U.S. demand (the IP-to-GDP ratio) has a more pronounced impact on Mexico than on China.

21. What may explain the insignificant role of the real exchange rate for Mexico? First, the share of peso-denominated inputs, that is, labor and services, has been comparatively small (Table 2). While labor, both skilled and unskilled, accounts for only about 15 percent of total cost in Mexico, its share in the United States and Europe is around 30 percent. Similarly, the share of service inputs is much smaller in Mexico. Second, the standard deviation of the real exchange rate has been small during the post-NAFTA period compared with the other variables in our sample.

Table 2.

Mexico, the United States, and Europe: Cost Structure in Export Industries, 1997

(In percent of total)

article image
Source: GTAP database.

22. To check the sensitivity of the results to changes in the trade regime, the VAR was re-estimated separately for 1980(1)-1994(4) and 1995(1)-2004(1) and we also checked the parameter stability in recursive regressions. The key link of the Mexico export share to the U.S. business cycle remains unchanged, albeit the absolute size of cumulative impact is larger in 1995-2004 as compared to 1980-94. Similarly, the direct relationship between Mexico and China exports remained statistically insignificant.

Displacement of Mexico’s Exports: Much Ado About Nothing?

23. Although the VAR results for aggregate series do not show displacement of Mexico’s exports by China, it is possible that such effects would be observed using disaggregated data. To this end, we compared detailed 2000-2003 trade series to see if losses in certain commodity groups by one country were offset by gains in other countries. As shown in Table 3, monthly changes in exports are positively correlated across countries. The relationships are qualitatively similar even when allowing for lags, different time periods, and alternative definitions of monthly changes.59 We also tested whether changes in China’s exports preceded changes in those of Mexico, that is, Granger causality, and we failed to detect such relationship. The only robust finding was Granger causality from U.S. industrial production or other measures of U.S. demand to sectoral imports. 60

Table 3.

Correlation Between Mexico’s and China’s Exports, 1997-2004

(Full sample, 2002:1-2004:4 in parentheses)

article image
Source: U.S. Department of Commerce; author’s calculations.

Additional Contributing Factors to Mexico’s Export Performance

24. In this section we briefly outline some additional, supply-side, factors mentioned in the literature that are difficult to capture in regression models.

25. First, the post-Tequila credit crunch has been relatively deep and protract, generating financial bottlenecks that have blocked growth in the tradable goods sector (Tornell etal., 2004). Indeed, real private sector credit declined cumulatively by almost 20 percent between 1999 and 2003. There are signs, however, that private sector credit has begun to pick up in the first half of 2004.

26. Second, the sluggish supply reaction of Mexico can perhaps reflect lower investment. While Mexico invests more than the average in Latin America, its total investment-to-GDP ratio and foreign direct investment are smaller than those of both industrial and transition countries. While it is too early to assess the impact of the current U.S. expansion on profitability and investment in Mexico, foreign direct investment is growing strongly and anecdotal evidence suggests that U.S. companies have resumed investing in manufacturing facilities in Mexico.

27. Third, Mexico has recently lost market share in one of its seemingly safest export industries—the automotive sector—because of its past orientation on, first, small- and medium-sized cars and, second, the Big Three producers (GM, Ford, and Chrysler). Both segments of the U.S. car market have declined recently. Regarding the former, Mexican factories specialized in the burgeoning SUV market have done well and are expanding. Regarding the latter, the authorities are trying to attract car producers from outside the United States to make automotive exports less dependent on U.S. producers.

28. Fourth, electric/electronic industries declined as a result of the product cycle: some goods have become “commoditized,” no longer requiring fast shipment or interaction with U.S. producers, and production has shifted to the country of the lowest cost producer. Finally, in surveys of investment climate, exporters and investors complain less about wage cost or exchange rates and more about issues such as a lack of skilled labor, poor infrastructure, red tape, and insufficient progress on structural reforms.61

29. Mexico’s proximity to the U.S. gives it a permanent advantage in industries with the following characteristics:

  • Product specifications are complex and routinely changing, and therefore requiring close interaction with the outsourcing company (electronic products in early stages of the product cycle, autoparts);

  • The inventory cycle for a product is short (all sectors where “just-in-time” delivery is required);

  • The goods are large in size, so that the cost of shipping offsets the cost advantages of producing in Asia (large household appliances).

Implications

30. The empirical findings suggest that Mexican exports exhibit a strong trend, affected by long-term outsourcing relationships with the U.S. firms, and are primarily driven by U.S. activity, while real exchange rate fluctuations have had a limited impact. To date, China’s exports do not seem to have displaced Mexican goods from the U.S. market, unless these exports had a clear cost advantage.

D. Effect of Trade Policy Changes and Other Shocks

31. To assess the implications of changes in external and domestic environment on Mexico exports, we employ the Global Trade Analysis Project (GTAP) model (International Monetary Fund, 2004). GTAP is a computable general equilibrium model that captures the geographic and sectoral structure of trade flows, providing a wealth of output at the aggregate and sectoral level. It is therefore well suited to analyzing the impact of free trade areas and changes in international and domestic economic conditions. For the purpose of this paper we focused on real GDP, welfare changes (equivalent variation of aggregate consumption),62 terms of trade, exports and imports, the trade balance, and sectoral output changes.

32. The standard GTAP framework, however, has limitations. First, being a comparative-static model, it calculates only changes to the levels of the relevant variables, not rates of growth. Second, it treats world economic growth as exogenous. Third, GTAP generally assumes that adjustment processes are smooth and costless. Fourth, the existing version of the GTAP database reflects data for 1997, that is, it reflects neither the long-run effects of NAFTA nor the entry of China into the World Trade Organization. Overall, the quantitative results obtained through the GTAP framework are sensitive to various assumptions, such as, for example, fixed versus flexible real wages, and the degree of international capital mobility. However, even with these caveats, the qualitative results provide insights into the relative importance of various shocks to the Mexican economy; the mechanism of structural and macroeconomic adjustment to these shocks; and the importance of the macroeconomic framework within which the economy reacts to the shocks.

33. The following shocks were analyzed: (i) introduction of CAFTA and the Mexico-Japan free-trade area, (ii) elimination of Multi Fibre Arrangement (MFA) quotas, (iii) lower transportation cost of Mexico exports to the United States, and (iv) a productivity increase in the Mexico manufacturing sector. We found the impact of various proposed trade arrangements to be broadly neutral, while the negative impact of MFA is more pronounced. Structural reforms that permanently increase Mexico’s productivity have strong positive effects.

Central American Free Trade Agreement (CAFTA)63

34. The first scenario evaluates the impact of the elimination of tariffs associated with CAFTA on trade flows among CAFTA countries (Table 4, first column). The aggregate impact is small, affecting primarily the clothing sector in Mexico (a cumulative decline of some 10 percent), while the impact on real GDP and aggregate welfare is negligible. Despite a deterioration in the terms of trade, a small depreciation of the real exchange rate leads to a positive impact on the trade balance, with significant resource reallocation away from the clothing sector and toward other manufacturing and service sectors.

Table 4.

Mexico: The Impact of Various Shocks on Key Macroeconomic Variables in the GTAP Model

article image
Source: Authors’ calculations.

Global investment is allocated evenly across regions.

Regional investment determined by the rate of return on investment.

Cumulative percent change.

Equivalent variation, in US$ billions.

Cumulative volume change, in percent.

In billions of US$.

Mexico-Japan Free Trade Agreement

35. The second scenario evaluates the impact of elimination of bilateral tariffs on non-farm products and a 50 percent cut in bilateral tariffs on farm goods between Mexico and Japan (Table 4, second column). The simulated outcome suggests a modest boost to Mexico-Japan trade and a small welfare increase for Mexican consumers, mostly through lower prices of imports. Under the assumption of a 50 percent tariff reduction, Mexico’s food exports to Japan would increase considerably, but the increases of other exports are limited because Japan’s tariffs on these products are already low. Mexico’s trade balance with Japan would deteriorate modestly, as domestic investment expands somewhat. These results are in line with the limited size of Japan’s trade with Mexico (less than 3 percent of total exports) and the offsetting nature of tariff cuts.

Multi Fibre Arrangement (MFA)64

The third scenario evaluates the impact of elimination of MFA quotas on exports and imports of textiles and clothing (Table 4, third column). This scenario results in somewhat larger GDP and welfare losses for Mexico than under the CAFTA scenario. A large contraction in textile and clothing output, by 14 percent and 35 percent, respectively, is offset by reallocation of resources toward other sectors, eventually resulting in a practically unchanged trade balance.65 Also, the regional composition of Mexico’s exports would change: losses in the U.S. and Central American markets are offset by gains in the markets of Europe and Japan.

36. It is interesting to compare the phase-out of MFA quotas with CAFTA tariffs, given that both shocks directly affect the textile and clothing industries, although the effect of the elimination of MFA quotas is narrower than CAFTA. It turns out that the former shock has a more pronounced impact compared with the latter, primarily because the export potential of CAFTA countries is limited and poses moderate competition in the U.S. market to Mexican exports, while the phase-out of MFA quotas will open further the U.S. market to exports from Asia.

Improvement in Mexico’s Manufacturing Productivity

37. The fourth scenario simulates the impact of raising Mexico’s manufacturing sector productivity relative to the rest of the world by 20 percent (Table 4, fourth column). Such an improvement could be brought about by reforms improving human capital and Mexico’s infrastructure, or by policies attracting more inflows of foreign direct investment into manufacturing, or both. This shock was simulated under two alternative assumptions about international capital mobility, first, with capital inflows allocated evenly across all regions (no disproportionate capital inflows into Mexico despite the productivity improvement), and second, with capital inflows determined by the expected rate of return on investment (strong additional capital inflows into Mexico).

38. While the two benchmark assumptions have essentially identical impacts on real GDP and welfare, they differ in their impact on the external balance. Real GDP expands in both cases by more than 12 percent and welfare improves dramatically, by the equivalent of about 10-11 percent of GDP. Gains in manufacturing productivity trigger an expansion in all sectors except food, mining, and services as resources reallocate toward textiles, clothing, and other manufacturing industries.

39. In the first scenario, controlling additional capital inflows narrows the investment-saving gap and keeps the demand for imports low. As a result, export growth outpaces that of imports by 9 percentage points, producing an improvement in the trade balance of more than 10 billion in 1997 US dollars or about 2½ percent of GDP, in spite of a deterioration in the terms of trade. This scenario illustrates the export potential of the Mexican economy, as long as the productivity gains are not offset by appreciation of the real exchange rate.

40. In the second, more realistic scenario, capital inflows increase sharply as a result of higher returns on investment, the domestic currency appreciates in real terms, and imports expand. Domestic saving only rise in line with GDP, however, leading to an increase in the investment-saving gap. The effect of increased capital inflows is reinforced by the rising prices of nontradable goods. As a result, imports grow faster than exports by about 15 percentage points, worsening the trade balance by about 14 billion in 1997 US dollars or about 3½ percent of GDP. This scenario illustrates the possibility of productivity gains in the tradable sector creating an investment and consumption boom and leading to a deterioration of the trade account.

Reductions in Mexico’s Transportation Costs

41. The fifth scenario evaluates the impact of improvements in Mexico’s infrastructure that would translate into a 20 percent reduction in transportation cost on Mexican exports to the United States (Table 4, fifth column). Overall, and contrary to anecdotal evidence, the impact is negligible, owing to the fact that transport costs for manufacturing exports to the United States account for just over 1 percent of the c.i.f. value, according to the GTAP database.66 A possible explanation for these counterintuitive results is that the GTAP database may underestimate the cost of transportation and the associated activities. The implicit costs of poor infrastructure, such as those resulting from long waiting lines at the border, or fraud at the customs, is large and left out from the GTAP calculations.

Implications

42. The GTAP simulations presented above assess the impact of various proposed trade arrangements and hypothetical changes in the performance of the Mexican economy. Regarding the former, we find that the Central America free trade area could have either a neutral or small negative impact on Mexico, in terms of GDP growth and external developments, while the Mexico-Japan trade agreement would have small positive effects; the impact of MFA is more pronounced, but still manageable, as long the economy exhibits flexibility to adjust to the external shocks. In contrast, should Mexico proceed with structural reforms that would permanently increase its productivity and rate of growth, the positive trade and welfare impacts are likely to be much larger.

E. Conclusions

43. The paper looks at factors that determine Mexico’s exports and that would explain the recent loss of U.S. market share. It finds that Mexico’s exports remain closely tied to U.S. activity, with little evidence of export displacement by China. The link to the U.S. business cycle is stronger than in Mexico’s competitors, as is the impact of changes in the composition of U.S. demand. Hence, the recent U.S. recovery, especially in industrial activity, may stop or perhaps even reverse the loss of Mexico’s export share. Simulations using the Global Trade Analysis Project model suggest that elimination of tariffs and quotas under various agreements (CAFTA, MFA, or Mexico-Japan free trade area) are likely to have a relatively small impact on Mexico. In contrast, structural reforms that substantially raise productivity in manufacturing could have a profound impact on exports and growth.

References

  • Acevedo, Ernesto Fernández, 2002, “Causa de la recesión en la industria maquiladora,” Momento Ecóndmico, No. 124, (Noviembre-Diciembre), pp. 11–25.

    • Search Google Scholar
    • Export Citation

  • Bergoeing, Raphael, Patrick J. Kehoe, Timothy J. Kehoe, and Raimundo Soto, 2002, “Decades Lost and Found: Mexico and Chile Since 1980,” Federal Reserve Bank of Minneapolis Quarterly Review, Vol. 26 No. 1, pp. 3–30. Also available at the internet: http://minneapolisfed.org/research/qr/qr2611.pdf.

    • Search Google Scholar
    • Export Citation

  • Konno, Toru, and Mototsugu Fukushige, 2003, “Did NAFTA Cause the Structural Changes in Bilateral Import Functions Between the U.S. and Mexico?” Journal of Policy Modeling, Vol. 25, pp. 53–59.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, 2004, World Economic Outlook: Advancing Structural Reforms, (Washington D.C.: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Kose, Ayhan, Guy M. Meredith, and Christopher M. Towe, 2004, “How Has NAFTA Affected the Mexican Economy? Review and Evidence,” IMF Working Paper 04/59 (Washington D.C.: International Monetary Fund). Also available at the internet: http://www.imf.org/external/pubs/ft/wp/2004/wp0459.pdf.

    • Search Google Scholar
    • Export Citation

  • Ocampo, José Antonio, 2004, “Latin America’s Growth and Equity Frustrations During Structural Reforms,” Journal of Economic Perspectives, Vol. 18, (Spring), pp. 67–88.

    • Search Google Scholar
    • Export Citation

  • Sims, Christopher A., J.H. Stock, and M. Watson, 1990, “Inference in Linear Time-Series Models with Some Unit Roots,” Econometrica, Vol. 58, pp. 113–144.

    • Search Google Scholar
    • Export Citation

  • Tornell, Aaron, Frank Westermann, and Lorenza Martinez, 2004, “NAFTA and Mexico’s Less-Than-Stellar Performance,” NBER Working Paper No. 10289 (Cambridge, MA: National Bureau of Economic Research). Also available at the internet: http://papers.nber.org/papers/w10289.pdf.

    • Search Google Scholar
    • Export Citation
1

Prepared by Aleš Bulíř. NitaThacker was involved in the design of the paper. Section D is based on inputs by Yongzheng Yang and Jean-Jacques Hallaert. Bruce Culmer provided splendid research assistance.

2

See Chapter I in this document. For a broader description of recent Latin American developments see Ocampo (2004).

3

Konno and Fukushige (2003) argued that the above-average export growth during 1995-96 was a cumulative result of liberalization processes that started in the 1980s, as opposed to a one-off impact of NAFTA. Another complicating factor is the effect of the sharp depreciation of the peso in late-1994 and 1995 on export performance.

4

A maquiladora is a Mexican corporation that operates under a maquila program approved by the Mexican Secretariat of Commerce and Industrial Development (SECOFI). A maquila program entitles the company, first, to foreign investment participation in the capital and management of up to 100 percent without special authorization; second, it entitles it to duty-free temporary imports, subject only to posting a bond guaranteeing that such goods will not remain in Mexico permanently. Formally, the current maquila legislation is the “Decree for Development and Operation of the Maquiladora Industry,” published by the Diario Official on December 22, 1989.

54

This comparison is based on the ratio of the value of 2002 and 2003 exports to an average of 1999 and 2000 exports.

55

Although during 1998-2003 the cost of shipping from China declined by about one-half in U.S. dollar terms, it still takes about three months for the goods to reach the U.S. customer.

56

An alternative to producer price indices used in industrial countries are indices based on unit labor costs. Unfortunately, consistent unit labor cost series are available neither for Mexico nor China.

57

Konno and Fukushige (2003) estimated bilateral Mexico-U.S. import functions with shift variables consistent with gradual switching.

58

See Kose et al. (2004) for a survey.

59

As export shares were found to be nonstationary, the series had to be differenced. The results do not change materially whether we use 12-month changes, month-on-month changes of seasonally adjusted (X12) series, or differences from the Hodrick-Prescott filter.

60

These finding are similar to those by Acevedo (2002), who investigated the causes of recession in the maquiladora sector in 2001-02. He found China’s products to be substitutes for maquiladora exports only in a few very narrow commodity groups.

62

Equivalent variation is the compensating payment that—in the absence of the economic change—moves the consumer to an identical welfare level as that associated with the change.

63

The free trade area includes the United States, El Salvador, Guatemala, Honduras, Nicaragua, Costa Rica, and the Dominican Republic.

64

Quotas introduced under the Multi Fibre Arrangement are being phased-out in stages over the period 1995-2005. The MFA (in place since 1974) resulted in a complex system of bilateral quotas on of textile and clothing exports from developing to industrial countries. For more detailed description of the MFA and Uruguay Round Agreement on Textiles and Clothing, which stipulate the phase-out of MFA quotas, see http://www.wto.org/english/thewto_e/whatis_e/eol/e/wto01/wto1_44.htm#note4.

65

The speed of such a massive reallocation is a feature of the standard GTAP model. It can be restrained by introducing a lower degree of factor mobility across industries.

66

The weights for food and mining exports are higher, but still less than 5 percent of the c.i.f. value.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Mexico: Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 2004: Issue 418
Publisher:
International Monetary Fund
ISBN:
9781451825626
ISSN:
1934-7685
Pages:
128
DOI:
https://doi.org/10.5089/9781451825626.002
  • View in gallery

    Mexico, China, and Canada: Export Developments, 1980-2003

    (In percent)

  • View in gallery

    Mexico: Nonoil Export Developments, 1994-2004 1/

  • View in gallery

    Mexico: Exports, the Real Effective Exchange Rates, and Prices, 1994-2004

  • View in gallery

    Mexico: Variables Used in the VAR Model, 1980-2004

    (Variables in first differences)

  • View in gallery

    Mexico, China, and Japan: Actual and Fitted Trade Shares, 1980-2004

    (In percent of U.S. imports, first differences)

  • View in gallery

    Impulse Responses of Mexico’s Export Share to Various Shocks

    (VAR in first differences with 3 lags; one standard error shocks)

  • View in gallery

    Impulse Responses of China’s Export Share to Various Shocks

    (VAR in first differences with 3 lags; one standard error shocks)

  • View in gallery

    Impulse Responses of Japan’s Export Share to Various Shocks

    (VAR in first differences with 3 lags; one standard error shocks)

Mexico, China, and Canada: Export Developments, 1980-2003

(In percent)

Mexico: Nonoil Export Developments, 1994-2004 1/

Mexico: Exports, the Real Effective Exchange Rates, and Prices, 1994-2004

Mexico: Variables Used in the VAR Model, 1980-2004

(Variables in first differences)

Mexico, China, and Japan: Actual and Fitted Trade Shares, 1980-2004

(In percent of U.S. imports, first differences)

Impulse Responses of Mexico’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)

Impulse Responses of China’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)

Impulse Responses of Japan’s Export Share to Various Shocks

(VAR in first differences with 3 lags; one standard error shocks)