A Panel Unit Root and Panel Cointegration Test of …

1 ECONOMIC GROWTH CENTERYALE Box 208629 New Haven, CT 06520-8269 ~egcenter/CENTER DISCUSSION PAPER NO. 942A Panel unit root and Panel Cointegration Test ofthe Complementarity Hypothesis in theMexican Case, 1960-2001 Miguel D. RamirezTrinity CollegeAugust 2006 Notes: Center Discussion Papers are preliminary materials circulated to stimulate discussionsand critical would like to thank Professor Gustav Ranis for reviewing this paper and Ms. LouiseDanishevsky for her editorial paper can be downloaded without charge from the Social Science Research Networkelectronic library at: index to papers in the Economic Growth Center Discussion Paper Series is located at: ~ Panel unit root and Panel Cointegration Test of the Complementarity Hypothesis in the Mexican Case, 1960-2001 Miguel D. RamirezAbstractUsing Panel data, this paper tests whether public and private capital have a positive andsignificant effect on aggregate output and labor productivity for Mexico during the 1960-2001period.

2 The richer information set made possible by the sectorial data enables this study toutilize the methodologically sound group-mean Fully Modified Ordinary Least Squares(FMOLS) procedure developed by Pedroni to generate consistent estimates of the relevant panelvariables in the cointegrated production (labor productivity) function. The results suggest that, inthe long run, changes in the stocks of public and private capital and the economically activepopulation (EAP) have a positive and economically significant effect on output ( and laborproductivity). The period is also broken down into two sub-periods: 1960-81 (state-ledindustrialization) and 1982-2001 (neoliberal model). The estimate for the public capital variablesclearly shows that it had a relatively more important economic effect during the earlier :Fully Modified Ordinary Least Squares (FMOLS), Panel unit Roots, Panel Cointegration Test, Complementarity Hypothesis, Mexican Labor Productivity JEL Codes: O10, O50, and O40 3I.

3 Introduction. Mexico s relatively successful state-led, inward-oriented strategy of economic growth anddevelopment was effectively brought to an end with the onset and aftermath of the debt crisis inthe early 1980s. In fact, it is easy to forget in this era of market-based and outward-orientedreforms that under state-led import substitution industrialization (ISI) Mexico enjoyed one of thehighest and sustained rates of economic growth per capita in the developing world--during theso-called miracle years of the late fifties and sixties. Yet, in less than two decades, Mexico scash-strapped governments have been forced to dismantle and privatize their massive state-owned sector, and following the country s accession to the GATT in 1986, its economy has beentransformed from a heavily protected and highly regulated one to, arguably, one of the most openand unregulated economies of the region. This market-led, outward-oriented process waseffectively locked in, both economically and institutionally, with the passage and phasedimplementation of the NAFTA beginning in 1994.

4 However, the country s transition from a closed to an open economy has been anything buteasy. It has often been marred by a series of economic crises and financial setbacks, mostdramatically in 1994-95 with the onset and aftermath of the so-called peso crisis , and mostrecently, with the severe economic slowdown the economy has experienced following therelatively mild recession in 2001. These economic crises and financial setbacks have led theMexican government to the involuntary adoption and implementation of several IMF-sponsoredadjustment programs which have resulted in sharp cuts in real government spending across theboard, the severe contraction and increase in the cost of credit, and last but not least, repeateddevaluations of the domestic currency in real terms [see Ibarra, 1996; Gonzalez, 2002; and4 Lustig 2001]. Not surprisingly, the economic growth and investment performance of the Mexican economyunder market-based reforms has fallen both far short of the initial (and inflated) expectations ofits more ideological advocates as well as when compared to its performance under state-led possible factor in explaining Mexico s poor growth and investment performance is the sharpfall in public capital spending demanded by the stringent fiscal deficit targets of the variousstabilization programs.

5 For example, Figure 1 below shows that overall public investmentspending as a proportion of GDP (RG) fell precipitously from percent in 1981 to barely in 2001; the dramatic fall in government investment is further revealed by the fact thataverage public investment spending as a proportion of GDP for the 1990s stood at just , which is less than half as much as the level recorded during both the 1980s and 1970s. From the standpoint of the composition of public spending, Figure 1 shows that publicinvestment channeled to industry (RGI) fell sharply after 1990 (from percent of GDP) to percent of GDP in 2001. On the other hand, state investment directed toward the servicesector (RGS) rose to over 2 percent of GDP after 1990 (as high as percent in 1991) andmaintained that level until 1994, after which it fell to levels slightly below those recorded in theindustrial sector (1 percent of GDP in 2001).

6 Finally, the share of public investment channeled tothe primary sector (RGP) has been historically low since 1970, and with the onset and aftermathof the debt crisis in 1982-83, it has declined to barely percent of GDP in In recent years, a number of investigators have undertaken (univariate) empirical studieswhich suggest that the dramatic fall in public capital investment experienced by developingcountries such as Mexico is of particular concern because investments in economic and social5infrastructure often generate substantial positive spillover benefits for the private sector byreducing the direct (and indirect) costs of producing, transporting, and delivering goods andservices to consumers [see Aschauer, 1989; Albala-Bertrand and Mamatzakis, 2001; Barro,1990; Cardoso, 1993; Green and Villanueva, 1991; Hermes and Lensink, 2004; Khan andReinhart, 1990; Moguillansky, 1996 ; Nazmi and Ramirez, 1997; Ram, 1996; Ramirez, 2002].

7 However, the major problem with several of these studies is that the data set available to test thecomplementarity hypothesis is often in annual terms and for a limited time period. Thus, evenwhen Cointegration tests are performed and error correction models are generated , the reliabilityof the estimates is questionable because unit root tests have low power when the number ofobservations is less than fifty as is often the case with univariate (annual) time-series In light of the above, this paper estimates a pooled model that attempts to determine whetherpublic capital in three major sectors of the Mexican economy has a positive and significant effecton Mexican output (and labor productivity) over the 1960-2001 period? The informationcontained in the time series data is thus enhanced by the cross-sectional (sectorial) data whichmakes it possible to reliably test whether increases in government investment spending enhanceoverall output and labor productivity in Mexico.

8 The focus on Mexico is particularly relevantbecause it is one of the few countries in Latin America and the Caribbean that has reliable anddisaggregated time-series data on public and private investment spending on a sectorial basisgoing as far back as the 1950s. More importantly, perhaps, it also allows policymakers todetermine where the effects of public investment spending, if any, are most significant. Andsince Mexico is a country faced with severe constraints in generating public revenues, anyadditional information that improves the allocation of scarce resources should prove highly6useful to the country s policymakers. The paper is organized as follows. Section II provides an economic rationale for includingthe public capital stock as an argument in a modified neoclassical production function, anddiscusses the empirical methodology to be employed in subsequent sections. Section III poolsdata for the primary (agricultural and mining), industrial (manufacturing), and service (banking,telecommunications, and energy) sectors and estimates a stacked production function (andlabor productivity function) over the 1961-2001 period.

9 This section also applies recentlydeveloped Panel unit root tests to the relevant variables to determine if they are stationary and apanel (and group) Cointegration test developed by Pedroni [1999a] is used to determine whetherthere is a stable long-term relationship among the relevant Panel regressors of the modifiedpooled production (labor productivity) function. In addition, it proceeds to estimate the pooledproduction (productivity) function via a group-mean Panel fully modified Ordinary LeastSquares (FMOLS) estimator developed by Pedroni [1999b; 2001] which not only generatesconsistent estimates of the parameters in relatively small samples, but also controls for potentialendogeneity of the regressors and serial correlation. This study thus represents an importantcontribution to the extant literature on the complementarity hypothesis because it addresses theimportant question of spurious correlation among the variables in the pooled (stacked) model.

10 The last section summarizes the chapter s major findings and offers some policy The Model and Econometric Methodology. Following the lead of Barth and Cordes [1980] and Aschauer [1989], it is possible toanalyze the potential impact of public capital spending on output and the marginal productivityof private capital in Mexico by appealing to the modified neoclassical production function given7in equation (1) below. Y = A F(L, Kp, Kg) (1) F1, F2 > 0; F11, F22 < 0; F12 > 0; F3 # 0; F23 # 0; F13 # 0. > > >where A is an index of multi-factor productivity; Y is the level of real output; L denotesemployment; Kp is the stock of private capital; and Kg refers to the public capital stock. By treating the public capital stock as a separate input in the production function, a ceterisparibus increase in public investment gives rise to three conceptually distinct First, if thepublic capital stock is productive and complements the private capital stock, a ceteris paribusincrease in the public capital stock will increase output directly in the same way that an increasein any other factor of production raises output (F3 > 0).