Trade and financial liberalization revisited: Mexico’s experience

Abstract

In the middle of two crises, the first one in 1982 and a second one in 1986, Mexico initiated a new economic strategy, inaugurated with two important reforms: trade liberalization, and financial liberalization and deregulation. With the first reform the economic authorities wanted to reposition Mexico in the international division of labor. With the second one they aimed at attracting capital inflows and modernizing the banking sector. Resolute commitment to drastic trade and to financial liberalization brought praise from a large part of the press and from international organizations, especially during the years from the mid-1980s to the mid-1990s when these two reforms were fully implemented. They were therefore greatly surprised when at the end of 1994 Mexico went into a deep crisis. The objective of this work is to study this period, to argue that the results of those reforms did not deserve to be praised, and to explain why the crisis could have been anticipated. Mexico’s experience is important in itself but we can also learn a lot from its positive results as well as its shortcomings. This is why an analytical reflection on the reasons for its failure to achieve the expected benefits is still important today.

Key words:

• general outlook and conditions
• Keynes
• Keynesian
• Post-Keynesian
• forecasting and simulation: models and applications

JEL Classifications:

• O54
• O19
• F41

Acknowledgements

The author would like to thank Carlos Tello, Guadalupe Mantey, Carlos Ibarra, Pablo Ruiz, Fidel Aroche, Tony Thirlwall, Robert Boyer, and Guadalupe Mantey for their comments to a previous version of this paper.

Notes

¹Figures for import permits do not extend beyond 1992, but at about that time trade liberalization had been completed. Note also, in this work I follow the Latin American convention that defines the real exchange rate Θ as N • (p^x/ p), where N is the nominal exchange rate (pesos per unit of foreign exchange), p^x is the index of foreign prices, and p is the index of domestic prices. Thus currency appreciation implies a falling real exchange rate.

²By the way, note that most of those changes took place before the North American Free Trade Agreement (NAFTA) came into effect in January 1994.

³Kalecki specified the wage-share equation using the concept of “degree of monopoly” rather than the profit margin. Since the two concepts are directly related, I prefer to use the latter, which has become more common in the contemporary literature.

⁴This hypothesis needs further study, but as a preliminary step I estimated a simple uniequational econometric model for private investment, where I included private investment I, GDP Y, the real exchange rate Θ, and import permits z (subindexes denote lags of the variables). The results were (variables are in logarithms):

$$i=3.57y_{-2}-2.54y_{-3}-0.43{\mathrm{\Theta }}_{-3}-0.17{z_{-1}}_{.}$$

Thus is appears that in the 1980–92 period import permits had a negative effect on investment, which were therefore favored by liberalization.

⁵The sign of the influence of z on I and on C^K may be different.

⁶The association between the wage share and z and Θ can be analyzed as follows. Consider Equation (2) ω = ω(λ, j, χ). The j variable, namely, the ratio of aggregate cost of materials to the wage bill, will grow with a higher rate of protection because this raises the cost of foreign inputs. This is also the case when the exchange rate is high. Besides, we can assume that λ, the average profit margin, also depends on the level of protection and on the level of the exchange rate: when either protection or the exchange rate is low, the domestic price of imported goods diminishes, and the pressure of foreign competition in the domestic market is correspondingly made stronger.

⁷Note, the concept of output at external equilibrium is simply the short-term version of Thirlwall’s rate of growth of domestic real income compatible with external equilibrium (Thirlwall, 1979).

⁸As shown, besides its short-term impact on exports and on the import coefficient, a change in the rate of protection has a long-run impact, because it affects investment and (probably also) the share of new investment channeled to the tradable goods sector.

⁹See, for example, López et al. (2011) and Caballero and López (2011).

¹⁰See Ibarra and Blecker (2013) for a justification for the inclusion of the two latter variables.

¹¹The 3.34 value of the parameter associated with foreign output y* may appear incredibly high. Note, however, that this parameter does not pertain to effective output y, but to output at external equilibrium y*. In Equation (6) the parameter for the former is estimated at 1.67. Moreover, it should be taken into account that U.S. GDP is several times larger than Mexico’s GDP so that a small growth of U.S. imports implies a very large rise in potential demand for Mexico’s exports.

¹²For more in-depth analysis using Mexico’s input–output matrix, see Aroche and Marquez (2012), Cardero and Aroche (2008), and Ruiz Nápoles (2001, 2007).

¹³A third factor that has conspired against substitution of imported inputs has to do with recurrent overvaluation of the domestic currency (see below).

¹⁴Note, however, that while a higher exchange rate would have favored both exports and import substitution, it would also have depressed aggregate demand and output (see Equation [6]). Certainly a competitive exchange rate, which is indispensable in any sustainable growth strategy, is not a cure-all measure.

¹⁵See especially Prebisch (1939 [1991] and 1944 [1991]). See also Mallorquín (2006) and Perez and Vernengo (2012).

¹⁶Prebisch did not use the term “financial fragility,” which is commonly associated with the name of Minsky (1982, 1986).

¹⁷This model is somewhat similar to the one proposed by López et al. (2006). I have omitted all variables of secondary importance, and B_Θ denotes the partial derivative of B with respect to Θ, and so on.

¹⁸In Prebisch’s story, bank lending is also stimulated when the rise in international reserves swells the monetary base. Mexico’s experience was different because the government sterilized a large parte of capital inflows. This is why international reserves R are not included as an argument in the C function. I owe this observation to Guadalupe Mántey.

¹⁹This type of development is beautifully analyzed in Ffrench-Davis (2010) and in Ocampo (2011), which is not surprising because these two authors have been prominent members of ECLAC, an institution that fortunately still carries the seal of Raúl Prebisch.

²⁰In this figure all variables have been adjusted to match mean and ranges. Therefore, the vertical axes have no precise meaning and have been omitted.

²¹From 1989 to 1994, domestic banking credit climbed from 27 percent of GDP to 52 percent. Concurrently, the share of households in total outstanding loans rose from 14 percent to 55 percent.

²²Let us recall here Kalecki’s formulation (Equation [1] above) in which profits are equal to the sum of capitalist expenditure in investment and consumption, plus the budget deficit, plus the foreign-trade surplus.

²³It is also useful to recall Keynes’s words: “Let finance be primarily national.”

Additional information

Notes on contributors

Julio López

Julio López is full professor at the Facultad de Economía of Universidad Nacional Autónoma de México.