Policy, Research, and Extemal Affairs WORKING PAPERS Agricultural Policy Agrculture and Rural Development Department and Latin America and the Caribbean Regional Office, Country Department II The World Bank February 1990 WPS 285 Enhancing the Contribution of Land Reform to Mexican Agricultural Development John Richard Heath Radical change in the land reform program is not in order in Mexico, but certain institutional changes would improve agri- cultural growth on farmlands governed by land reform. The Policy, Rarch. aid Extemal Affails Coanpilx distribute PRE Woating Papers to disneminue the findinW of wock in p and to ricourage the exdcange ofideas anong Bank gaff an all othr interested in developnat isue. Thaes papas crry the nanes of the authei, reflct only thir views, and should be used and cited accowdingly. Tlhe finding, inuapreationa. and caiclusiona ar the Polcy R-arch, an Exealwm Afar G = _~le Agricultural Policy This paper - a joint product of the Agricultural Policy Division, Agriculture and Rural Development Department, and the Agriculture Operations Division, Latin America and the Caribbean Regional Office, Country Department II - assesses the institutional aspects of agricultural development. Reforms in land and credit policies, which are the focus of this effort, can have a major impact on both equity and agricul- tural growth. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washing- ton DC 20433. Please contact Cicely Spooner, room N8-039, extension 30464 (85 pages with tables). The ejido is a rural community on which the * Simplify and clarify restrictions for private Mexican government has conferred land and farmers on holding size and land use. water resources. Ejido members (ejidatarios) can use the land but are prevented by agrarian * End restrictions on renting or sharecropping reform law from selling it. The ejido seems to by ejidatarios. be a more or less fixed element in the Mexican rural economy. * Allow ejidatarios to sell their land parcels to other members of their ejido (not outsiders). Heath found no conclusive evidence that individual. ejidos are significantly less produc- * Improve management of communal lands. tive than private farms, and hence it seems unlikely that privatization of ejidos would * Extend credit directly to individual ejidatar- greatly improve agricultural growth. At the ios, on the basis of their creditworthiness. margin, however, ejidatarios face more con- straints on productivity growth than private * Cease having the whole ejido bear the farmers. burden of loan default by one or more ejidatar- ios. Heath recommends the following piecemeal improvements to the existing structure: * Provide credit to ejidatarios wholly in cash and allow them to decide what inputs to pur- * Accelerate the drive to give ejidatarios chase and what crops to plant. titles to their parcels of land. * Grant ejidatarios titles irrespective of the size of their parcels. The PRE Working Paper Series disseminates the findings of work under way in the Bank's Policy, Research, and External Affairs Complex. An objective of the series is to get these fmdings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in these papers do not necessarily represent official policy of the Bank. Produced at the PRE Dissemination Center TABLE OF CONTENTS Page I. Executive Summary ......... ............ 3 II. introducton ........................... 9 III. Historical Background ........ ........... 11 IV. Problems with the Legal and Administrative Framework of the Land Reform ...... ....... 18 V. Access to Land Use and Land Use Patterns .24 Vt. Labor Organization and Employment .29 Vii. Access to Credit .35 Vill. The Ejido Marketing Support Program .41 IX. Comparing the Performance of Ejldos and Private Farms .44 X. Conclusions and Policy Recommendations .55 Tables .59 Appendix A: Interpreting the Results of the 1981 Agricultural Census .80 Blbiography .82 The author woula like to thank GIOIA PALMIERI for her help as a research assistant. 1. EXECUTIVE SUMMARY Background in Mexico, the cohmnitment to and reform Is enshrined kI the 1917 Constitution and the reform remains, to this day, an kmportant politieal symbol By 1988, no less than 54 percent of national territory had been affected by the land reform. The land reform peaked In Intensity during the C6rdenas administration (135-40) and, since the mid-1970s, there has beon a significant drop In the amount of land redistrlbuted. Slnce 1977 governments have often stated that there Is no more lnd available to redistrlbute; emphasis has shifted from redlstrbutlon to ensurig that landholders (both within and outside the ejido sector) have adequate legal title to their land. Although expansion of the land reform sector Is not part of the present government's policy agenda, It derives a significant proportion of Its electoral support from the ejido population. Some have argued that the ejido system has always been of greater Importance as a polltical control mechanism than as a tool of economic policy. In any event, the ejldo's political signlficance should not be underestimated. Approach Consideration focusses on Individual ejldos since these constitute the majority; collective ejidos are few In number and, since the mid-1 970s, no govemment has promoted collectivizatlon. An attempt Is made to compare the performance of Individual farms Inside the ejido with comparably-sized private farms. in Mexico, large farm enterprises have traditionally been favoured by government programmes of subsidized Input delivery. Therefore, In reviewing the historical evidence concerning the relative efficiency of ejidos and private farms It Is important to control for the scale of enterprise. Wherever possible a comparison Is made between ejido parcels of up to 5 hectares and private farms In the same size range. According to the latest agricultural census data (1981), roughly 60 percent of the farms In both tenure categories are of this size. Of the land In holdings up to five hectares, 75 pereont Is controlled by ejidos and 25 percent by private farms; the average size of farms In the private sector Is 74 hectares, compared to 7 hectares In the ejido sector. Data Limitations Comparisons of the performance of both tenure categories are hampered by the Incompleteness of the 1981 census data; In many cases (e.g. when comparing crop yields) the most recent data derive from the 1970 census. Therefore, It Is impossible to draw hard and fast conclusions about the relative productivity of private farms and ejidos in the recent period. However, by observing differences In the Institutional arrangements confronting Individual ejidos and private farms It Is possible to Infer whether such differences are likely to lead to variations In productivity between the two tenure categories. 4 Problems with the Land Reform Law Vagueness and contradictions In the law create a climate of uncertainty that may discourage on-farm investment by both ejidatarios and private farmers. First, whil there Is some logic to estabiishing lknits to the size of holdings in the private farm sector, the ratinale for varying these cellings according to the type of crop cultivated Is dubious. Also, In the case of cattle raising, the law Is Inconsistent about the maxinum holding size permitted and there Is a legal ioophole (amparo) that has resulted In the selective enforcement of these cellings, undermining the credibillty of the law. Moreover, It Is unclear what proportin of their land cattle-raisers may legitknately devote to crop cultivation - The legal stipulation about the size of parcels In individual ejldos (20 hectares of rainfed land) Is not enforceable owing to demographic pressure; in some cases, ejidatarios have falled to recelve title to their holding because the parcels Is smaller than the officlally-prescrlbed minimum. Although access to state credit appears not to be contingent on possession of title to the land, the lack of a titulo parcelarlo does reduce the ejldatario's security of tenure: It may Increase the likelihood of his being arbitrarily evicted If he falls out with ejido authorities. Probably the most Important area of uncertainty concerns the legitimacy of renting-out ejldo parcels (or engaging In sharecropping). While Article 76 of the 1971 Ley Federal de Reforma Agrarla may be interpreted to permit renting In certain circumstances, there Is a noticeable lack of agreement about the legality of this practice. The 1981 Ley de Fomento Agropecuarlo does not eblfnate the confusion: while facilitating "iolnt ventures" between ejidatarios ard private farmers It Is not clear from this law whether forms of association based on renting are Illegal. Access to Land and Land Use The fragmentation of holdings Is only slightly less advanced in the ejido sector than among small private farms Indicating that, despite formal prohibitions on the division of ejido parcels on Inheritance, the land reform has largely failed to check the spread of minifundia: on holdings up to five hectares, the average size of farms was 2.4 has, In the ejido sector and 1.7 has. n the private sector (1981). Also, In 1981, In the case of these small farms, there was no significant difference between tenure categories In the crop mix, in the proportion of cropland left In fallow and In the number of cattle per hectare. However, there were two significant differences between the tenure categories. First, Irrespective of holding-size, the ejidos reported a higher Intensity of land use In 1981 (proportionately more land In agricultural production). This may be partly attributable to the second factor: ejidos had better access to irrigation than small private farms (an apparent reversal of the pre-1970 trend). Although the exact proportion Is hard to determine, a large part of the land in Individual ejidos is reserved for communal use: primarily as forest or rough grazing. Slnce no Individual ejidatarlo feels responsible for the communal reserve 5 this may encourage overgrazing and poor forest management; If widespread, these practices will reduce the average productivity of the ejido sector In relation to private farms. Employment Owing partly to holding fragmentation, small farmers and ejidatarlos are not able to depend entirely on the output of their land to cover family subsistence requirements: farmers in both sectors are heavilly Involved In off-farm work, seeking employment In the large towns and also In the United States. Recourse to off-farm work has been further encouraged by urban real wage trends which (until recently at least) have been more favourable than trends In the farm-gate prices of staple crops. Rural outmigratlon has generally caused (and been a consequence of) trends In agriculture Involving less Intensive use of labour per hectare. Since the 19709, these trends have centred on more widespread use of tractors, an Increase In the relative Importance of livestock Income and a shift toward less labour-intensive crops (notably sorghum). For ejidatarlos, the Increased dependence on off-farm income sources appears to have led to greater recourse to leasing-out of ejido parcels and (equally Important) to sharecropping. As well as placing ejido land In the hands of private farmers, these practices also create employment within the ejido (since land-deficit families that do not migrate are able to work the land of those who have migrated): although, In many cases, renting and sharecropping Increase wealth concentration, they may In some cases have positive distrlbutional effects; In any event, they ensure efficient exploltation of ejido land by placing it In the hands of those with the means and the vocatlon to work It. Credit EJldatarlos have less access than small private farmers to commercial loans because the law prevents them from mortgaging their land; there Is therefore a long traditlon of dependence on state credit banks that have waived the collateral requirement and, traditionally, exercised weak sanctions against loan default. There is a strong ethos of paternalism and low credit discipline In relatlon to development bank lending to agriculture. This Is a significant point, first, because the ethos Is pervasive, affecting development-bank (Banrural) lending to private farmers as well as ejidatarlos; second, because It Is associated with the tendency to use credit for political rather than economic purposes. The second of these factors contributes to a tendency to spread credit resources too thinly: partly in order to enhance rural tranquility and partly from a misguided attempt to use working-capital loans as a means of contributing to broad-based social welfare. Patemallsm Is discernible In the long tradition of restricting the range of crops for which state credit Is available (In effect, telilng farmers what they may grow); also, In the continuing provision of credit In kind, with the bank assuming responsibility for selection and timely delivery of Inputs. These practices respectively discourage crop diversification (potentlally restricting farm Incomes) and adversely affect crop yields by giving the farmer no freedom to tailor Input application to his precise requirements. 6 Although, In these respects, ejidatarios and private farmers lending from Banrural face the same constraint, there are two obstacles to credit access that apply specifically to ejidatarios. First, IndIvIduals within the ejido cannot negotiate directly with Banrural for their loans: the bank Is legally required to operate with the ejido as a whole through the ejido leadership; individual credit appilcatons must utimately be endorsed by that leadership. Therefore, ejidataris may potentially be exckided from credit If they have political differences wlth the leadership. Second, In the event of default by some of the ejidatarlos who have received Banrural credit, the whol ejido may be cut off from credit In the subsequent crop cycles: ejHdatarlos who repay their debts lose out to the Irolvency of their feolws. Since ejldos tend to be large In size and diffuse In membership characteristics, It Is rarely the case that there Is a -cohesive peer pressure group capable of ensuring that those tempted to default honour their debts. The probabillty that the ejido will be cut off from credit In the following cycle (coupled wlth the bank's tolerance of Individual defaults) serves to undermine credit discipline. Marketing A programme exists that subsidizes the transport costs of ejidatarlos and small private farmers who sell grain to the rural warehouses operated by the state. While this programme may marginally extend the reach of the government's guarantee price policy, its overall Impact is restricted by the nature of the warehouse network: warehouses are concentrated In better-off regions (where the bulk of the grain surplus Is produced) and there are few of them In poor, remoter regions (where the need for Income support Is greatest). Therefore, marketing subsidles (Ike credit subsidies) have leaked" to better-off farmers, wasting resources. Comparing the performance of ejidos and private farms From the preceding discussion, four hypothese may derived, all tending to suggest that ejidos should be less productive than private farms: 1. Many ejidatarlos lack usufruct title to their land, reducing their security of tenure and inhibiting on-farm investment In the ejido sector; 2. Unclear laws concerning renting and sharecropping Inhibit the transfer of land use to the most officient producers; 3. Mismanagement of ejido communal land depresses the average productivity of this sector In relation to private farms; 4. Problematic access of ejidatarlos to credit adversely affects crop yields. On the other hand, there Is a countervailing hypothesis: 5. Contradictory laws concerning holding-size likits and land-use restrictions may Inhibit on-farm Investment by private farmers. 7 Testing of these hypotheses Is difficult owing to data lnitatlons. There are no recent studies of the relative productivity of ejidos and private farms: the avalable sources are based either on agricultural census data up to 1970 or pre- 1970 case studies. In the period before 1970, ejidos appear to have had poorer and on average than small private farms and to have had lesser access to credit and public investment; also, there Is evidence of scale-blas In the government's dolvery of subsidized capital Inputs. These factors should be borne In mind In interpreting pre-1970 studies. Sources based on the census record reach varying conchjsions about the relative productivity of the two tenure sectors- summarizing, It wou appear that - ejidos had lower land yields before 1970 but higher output per unit of purchased capital kiputs; ejldos also show lower labour productivity than small private farms, al the more so If a cost Is Imputed to unpaid family labour. However, given the clndestine renting out of ejido land to private farmers, the census data probably fall to reflect the real distribution of Inputs and outputs between the two tenure sectors; this places In doubt any concluslons about relative productivity that are derived from the census data. Of the case studies, only one (dating to 1967) controls effectively for both land tenure and enterprise scale: the study concludes that, In Irrigated reglons, ejidos outperform private farms in each size stratum (the gap between sectors being largest In the case of capital productivity, confirming the results of census-based studies). The Incompleteness of the 1981 census data does not permit calculatlon of relative productivity In the recent period. However, land use data from this census suggest that ejidos are likely to be at least as productive as small private farms: ejldo parcels up to five hectares are better Irrigated than comparably- sized private farms and show a higher proportion of land in agricultural use; also, the ratio of fallow land to cropland Is the same for both tenure categorles In this size class. Thus, controlling for enterprise size, there Is no conclusive evidence that Individual ejidos are less productive than private farms. Policy recommendations There are major political constraints to tampering with the Mexican land reform; moreover, the evidence does not suggest that radical change In land tenure would greatly enhance agricultural growth. However, while the Institutional differences between individual ejidos and small private farms appear not to generate major ditfferences In Indicators related to productivity, at the margin, ejldatarios face more constraints on productivity growth than private farmers. Rather than seeking to undo the land reform, policy Initiatives should focus on making piecemeal knprovements to the existing structure. Specifically, 1. Granting of titles to ejidatarlos should not be contingent on parcel size. 2. The existing drive to provide ejidatarlos with usufruct titles should be accelerated. 8 3. Holding-size and land-use restrictlons for private farmers should be skipilf led and clarified. 4. There should be no restrictlon on renting or sharecropping by ejldatarlos. S. Ejldatarbos should be able to sell their parcels to other members of their ejido (not to outsiders). 6. Management of communal lands should be knproved. 7. Credit should be extended directly to Individual ejidatarios, on the basis of their creditworthiness. 8. The burden of loan default should not be borne by the whole ejldo. 9. Credit should be provided wholly In cash and ejidatarlos should be free to purchase the Inputs and plant the crops they wish. 9 II. INTRODUCTION 1. The purpose of this paper Is to conskier whether the hstitutional factors associated with the Mexican land reform (1Jkb) sector constrain agrkcultural growth. Parthg from the assumption that the ejido may be considered a more or less fixed eement of Mexico's cuWtural and politial scene, the paper wUI focus on the pecomeal changes that could help to enhance growth and productivity within the existing structure. 2. The ejldo Is a rural community, possessing land and water resources conferred upon it by the Mexican government; ejido members (ejldatarlos) have usufruct rights over the land but are prevented by the agrarian reform law from selling it. The term "ejido" originally referred to communal lands outside (at the "exit") of the medieval Spanish village; as an institution, the ejldo also has antecedents In pro-Columblan land-holding patterns (Eckstein,1966:9-13). 3. There is little rellable, recent Information on the extent of the ejido system (see Appendix A). In 1970 (the last date for which systematic farm census data Is available), about 60 percent of farm families were elidatarlos (Yates,1 981 a: 1 45). At that tkie there were over 23,000 ejidos In Mexico, possessing 56 percent of arable land and accounting for almost 40 percent of the combined value of crop and livestock output (Yates,1981a:160). 4. In considering the constraints posed by the ejido sector It Is necessary to ask whether the institutional differences between ejldo enterprlses and skmilarly-sized private farms give rise to significant variations in the economic performance of enterprises In each tenure category. To enhance the legitimacy of the comparison, references to the ejido sector will focus on Individual ejldos: 10 those communttles In which ejido households are each allocated their own parcel of land, this land being worked on an Individual rather than a collective basis. An ejldo of this typo Is offectIvely a nucleus of separate household-based enterprlses; in terms of production and marketing procedures, It has much In common with villages whose inhabitants are small private landowners. 5. In 1981, 58 percent of the parcels In Individual ojldos occupied an area of five hoctaros or ess; the average size of each parcel was seven hectars. hI the- private farm sector, 56 percent of enterprises covered no more than fIve hectares while the average farm size was 74 hectares (Table 1). Thus, although there lb roughly the same proportlon of small enterprises In both sectors, large private farms are much bigger on average (In terms of area occupied) than the largest parcels In the ejldo sector: most of the giant cattle ranches are private farms, not ejidos. 6. There are a small number of collective ejidos where the land Is worked In large, centrally-managed units with each ejido household being assigned a work quota. The Echeverria government (1970-76) created about 5,000 collective ejidos but, partly because these were "legislated Into existence from above", they met with considerable peasant resistance (Barkin,1977). There is no reliable data on the number of collectIve ejidos currently In operation (or the area they occupy); In 1970, they accounted for less than 5 percent of the arable land In the ejldo sector (Eckstein et al,1978:36). 7. The collective ejldos are concentrated In the Laguna region (which overlaps parts of the states of Coahulla and Durango); the Yaqul Valley (Sonora); Los Mochis (Slnaloa); and the state of YucatAn (Johnston,1983:235n). It Is safe to conclude that, at the moment, the collective ejidos do not occupy a promInent place on the Mexican rural scene: since the Echeverria administration, the MexIcan 11 government has shown no commitment to the principle of collectivization (Montanez,1989). Therefore, for the purposes of this report, It Is not biportant that they be excluded from the analysis. 8. The land reform sector (the so-called sector social) consists not only of ejldos but also Indigenous communities: the latter are mainly settlements of Pre- Cohmblan origin and are less numerous than the ajldos. This report will not -comment on the speclfic constraints facing the indigenous communKtles. Ill. Historical Background 9. Throughout the colonial and postcolonlal period of Mexican history, there was a high level of concentration In the pattern of landholding; following the liberal reforms of the mid-nineteenth century (which abolished the Inalienability of church property), the process of concentration accelerated, receiving much stknulus from the expansion of commercial agriculture that occurred under the rule of Porfirlo Diaz (1876-1910). During the Porfirlato peasant villages lost much of thelr land to the expanding estate sector and by 1910 the degree of land concentration In Mexico exceeded that of any other Latin American country (Ecksteln et al,1978:17). 10. Land reform was called for by one of the principal factions contending for power In the Mexican Revolution (1910-20): the peasant Insurgents from the state of Morelos, led by Zapata (Womack,1969). Ultinately, however, It was not the zapatistas but liberal-bourgeois elements that seized control of the Revolution. While this faction fully appreciated the kmportance of ensuring peasant support they remained largely unconvinced that agrarian reform might serve as a sultable foundation for agricultural growth. During the period of regkne formation (1910- 40), from all sides of the political spectrum, "the peasantry constituted a vital polltical force"; "All major groups contending for national power therefore promised 12 to return communal lands to their traditlonal owners, less out of conviction, In some cases, than out of political necessity" (Hewitt de Alcintara,1980:23). 11. The founders of the modern state apparatus (notably Calles) halled from northern Mexico and were strongly infiuenced by the U.S. pattern of agricultural development: they felt that large commercial farms, private enterprise and Increasing capital Inputs were necessary for agricultural growth, remaining unconvinced that cooperative organizations of small peasant producers coUld either gonerate the marketed surplus of staple foods needed for urban industrialization or effectively sponsor agricultural export growth. In the 1920s, President Calles envisaged that the ejido parcel would never constitute more than a partial subsistence base for a rural population primarily dependent for Its llvelihood on wage earnings: the parcel was Intended to provide an Income supplement rather than constitute a viable farm exterprise (Hansen,1971:107-108; Johnston,1983:234; Hewitt de Alcantara,1980). 12. Although the commitment to land reform was enshrined In the 1917 Constitutlon, significant redistribution of land was delayed until the mid-1930s. During the CArdenas administration (1935-40) the land reform sector more than doubled in size (Table 1). However, It has been argued that the government's objectives were more political than economic: land reform was necessary to break the stranglehold of regional oligarchles and to facilltate the centralization of power, while simultaneously cementing the loyaltles of the peasantry to the PRI regkne (Hansen,1971; C6rdova,1974; Rello,1987). Therefore, the reform was carried out hurriedly and unsystematically wlth little attention paid to the quality of the land endowment or the vlability of the size of holdings created (Brannon & Bakianoff,1984:1133). 13 13. However, despite the relative intensity of land reform In the mid-19309, a significant number of the large estates were left Intact. Guarantees against expropriation were extended to many private cattle producers. Also, It may be argued that cardenista policy went against the grain. Both before and after the Cirdenas administration skepticism toward land reform dominated thinking In the upper echeons of government: the spirit of Calles ultimately proved more enduring than the spirit of Cirdenas (Hanson,1971). 14. After 1940, there was a progressive reconsolidation of private property, much of It by persons well-connected wlth the PRI regime (newcomers to the countryside who were scathingly referred to In Mexico as agricultores nylon (INEGI,1986:268-268D). This was associated wlth the formation of neolatifundla, private properties exceeding the holding-size lkmit prescribed by land reform: these properties would be nominally parcelled out among several different owners (usually relatives of business partners), but would In practice be operated as a single unit (Stavenhagen et al,1988; Warman,1977). 15. The commitment to land reform was vitlated, not only by the survival of large private estates, but more significantly by the government's orientation of public Investment and agricultural research toward the needs of large private farms rather than ejidos. Johnston and Kllby (1975) have argued that, in contrast to Taiwan and Japan (where the land reform was more thorough going), In Mexico, the coexistence of large and small farm units was accompanied by a government pollcy that gave preference to the former: the Incompleteness of the agrarian reform was compounded by a mscale bias" In Input delivery. The private farms benefitted particularly from the huge, reservoir-based irrigatlon projects that were set up In the north and north-west regions, regions that became the prime 14 contributors to the rapid agricultural growth of the 19409 and 19509 (Hewitt de Alcintara,1978; Wlonczek,1982). 18. Although the large collective ejidos that had been established by CArdenas continued to absorb significant credit and Investment resources the bulk of the land reform sector (consisting overwhelmingly of Individual ejidos) received only marginal attention In the period up to 1970. The sector was starved of Institutional credit: wThe total number of land reform beneficiaries with whom the EJldo Bank worked In any given year between 1950 and 1970 averaged only 32,000, or roughly two percent of all the ejidatarlos In the country" (Hewitt de AlcAntara,1980:30). 17. The stagnation of the small-farm sector (private farms as well as Individual ejidos) carried a significant cost. By the early 1970s the agricultural product had begun to grow more slowly than Mexico's population; food Imports started to rise sharply (Heath,1988; Yrnez Naude,1988). Slmultaneously, there was an outbreak of widespread rural unrest, Involving the formation of peasant organizations independent of the Natlonal Peasant Confederatlon (CNC), a ruling-party organ that served as one of the regime's support mechanisms. The Echeverria government (1971-76) responded to the twin threats of agricultural stagnation and political agitatlon by significantly Increasing the volume of public resources channeled toward smaliholder agriculture, particularly In hitherto-neglected rainfed areas. The private, irrigated farms of the north continued to soak up a share of credit and Investment more than proportional to their contribution to crop output value; but, In relative terms, central and southern regions (where ralnfed agriculture prevailed) fared better than they had done before 1970 (Goodman et al,1985). 15 18. Increased pubilc spending on agriculture In the 1970s (Cumilnating in the ambitkus SAM programme of 1980-82) led to some knprovement In crop output growth. Taking three-year averages (to reduce the hnpact of rainfall variatlons between crop years), and comparing equdistant periods, real crop GDP grew by only 1 1 percent between 1967-69 and 1973-75; however, from 1973-75 to 1979- 81, growth more than doubled, rising to 25 percent (World Bank,1989:74). However, the growth-iducing kmpact of goverrvnent Input subsidies was partlfay undercut by a pricing policy that discrkninated against the rural sector. Two kmportant studies have Independently reached a significant conclusion about 1970-'82 trends: aithough total Inter-sectoral resource flows (transfers through the price mechanism added to fiscal transfers and transfers via the banking system) showed a net gain for the countryside, relative price movements continued to favour the urban rather than the rural sector (G6mez Oilver,1984;Goodman et al,1985:Appendix A). 19. In the late 1970s and early 1980s public spending on the agricultural sector was fuelled significantly by the expansion of oil export revenues and overseas borrowing, a policy that was not sustainable after 1982. Significant cutbacks In Input subsidies contributed to a slowing of agricultural growth, although the growth rate remained somewhat higher than It was In the early 1970s: between 1979-81 and 1985-87, crop GDP increased by 13 percent In real terms (World Bank,1989:74). Post-1985 changes In exchange rate and pricing policy may ultinately stinulate agricultural growth, more than off-setting the Impact of cuts In Input subsidles: It Is too early to draw any concluslons In this respect. 20. What impact have these recent sectoral trends had on the ejido? After 1970, a larger share of pubiic resowrces were directed toward the small-farm sector and toward rain-fed agriculture. For example, irrigatlon schemes became 16 smaller In scale and altered their regional focus; several ejidos In central Mexico benefitted from tubewell projects. Essentially, ejidos benefitted Incidentally from a policy switch whose prkme purpose was to redress the stagnation of rainfed agriculture In general. The one ejido-specific policy of note was a botched attempt at collectivizatlon; this met with widespread peasant resistance and was abandoned with the change of administration In 1978. Both during the oil boom and In the post-1982 period of fiscal austerity, agricultural development initiatives do- not appear to have discriminated significantly between ejidos and small private farms. 21. While, In principle, the Mexican government has continued to support the agrarian reform, the impact of land redistribution has dropped sharply in recent decades. in terms of hectares distributed, the most active periods of land redistribution were 1935-40 and 1984-70 (Table 1): more land was redistributed In the second of these periods but little of It was prime cropland and, since the settlement density was much lower, the number of reform beneficiaries was less In 1964-70 than In 1935-40 (respectively, 278,214 and 728,847). Slnce 1976 the pace of land reform has slowed considerably. In 1977-88, the average area of land formally redistributed each year was only 0.8 million hectares compared to 3.1 millon hectares per year In the previous twelve year period (1965-76); the mean number of beneficiaries per year fell from 40,351 In 1965-76 to 36,283 In 1977- 84 (INEGI,1986:273). 22. On various occasions during the past two administrations government ministers proclaimed that there was no more land left to redistribute (Baaley,1981:358; Heath,1985:115). Accordingly, the focus of the agrarlan reform has shifted away from handing out land toward a policy of encouraging private sector Investment In the ejido sector and "regularizing" the tenure status of all farmers by ensuring they have adequate title to their land (Heath,1985). 17 23. What Is the current extent of the land reform sector? By 1988, no less than 54 percent of national territory had been affected by the land refom (Table 1). However, only 12-13 percent of Mexico's territory (area: 197 mllion hectares) was under cultivation In this year; moreover, It Is estknated that the maxhwum feasible area for cropping Is no more than 35 millon hectares or 18 percent of the total land area (Yates,1981a). Thus, even If all the land suitable for cropping were put under cultivation, more than half of the land In the ejido-sector would remain In forests, rough grazing or shuply devoid of productive use: this represents an Important constraint on agricultural growth and employment generatlon within the ejidos. 24. Focussing on land redistributed as a proportion of the total area of each region, the South Pacific and Gulf are the regions most affected by the redistributlon: In each case, 52 percent of the regional area had been subject to reform by 1985 (Tables 2 & 3). The least-affected region (with 45 percent of the land area In the reform sector) Is the North Pacific: this Is partly because, given the sparsity of settlement here, there have been few Invasions of estates by the landless. Also, In the Immediate aftermath of the Revolutlon, many ruling-party politicians either owned or proceeded to acquire large tracts of land In this wealthy agricultural reglon. Some authors maintain that there was a certain resistance within the PRI regime to carrying out land reform In the North Pacific (Hansen,1 971; Johnston,1983:235). Writing of the so-called "Sonora dynasty", Sanderson notes that "revolutionary generals in the North, once in power, had llttle Interest In distributing the new wealth they acquired" (1984:45). 18 IV. PROBLEMS WITH THE LEGAL AND ADMINISTRATIVE FRAMEWORK OF THE LAND REFORM 25. Rather than provido an account of the legal basis of the Mexcan land reform (which has changed significantly since 1917), or a step-by-step description of the administrative procoedur by which ejldos are created, this section will focus on some of the main problems created by the framework as It exists today. A detailed account of the legal and administrative framework and how It evolved may be found In Craig (1983), Diaz Cisneros (1983) Sanderson (1984) and Zaragoza and Maclas (1980). 26. Two points should be borne In mind. First, the legal foundation of the Mexican land reform derived Initially from Article 27 of the 1917 Constitution which affirms that all land and subsoil resources are ultimately the property of the nation. While the principle of private property ownership Is conceded the state reserves unto Itself the right to modify the existing tenure pattern In order to ensure a more equltable distributlon of wealth; to this end the state Is authorized to expropriate private property (Zaragoza & Maclas,1980:53-54). 27. Second, according to the law, the ejido is "an Institutlon, formally constituted by the federal government, by which property Is assigned to a given population group; the ejido is a form of social property; property rights conferred on this population group are Inalienable, non-transforable and may not be confiscated; the property shall be exploited as an Integrated unit of production, preferably organized along collective lines; It shall be endowed with executive powers and function according to the principles of Internal democracy, cooperation and self-management" (Zaragoza & Maclas,1980:207). 19 28. Since the rate of creation of now ejidos has slowed significantly since 1976 and Is unikely (owing to the unavallability of approprlate land) to pick up in the future, discussion will not focus on the problems associated with the procedure for granting land. Instead, It will consider two important respects In which the difficultlis of Interpreting land reform legislation constitute a source of insecurity for exbting ejidatarlos and private farmers. (a) Holding-size speciflcations and land titles. The agrarian reform statutes lay down holding-size limits for ejpdos and private farms. Limits on the size of private farms take Into account variations In land productivity: one hectare of Irrigated land Is assumed equal to two hectares of rainfed land, four hectares of pasture land and eight hectares of brush or rough grazing In arid lands. Private farmers may own a maxinum area of 100 hectares of irrigated land, or the equlvalent In other types of land. However, there are several exceptions to this rule, exceptions whose wisdom has been disputed (e.g. by Yates,1981a). 29. in the case of cotton, properties of up to 150 hectares of irrlgated land Cor the equivalent) are allowed. The likit Is set at 300 hectares (irrigated) If the crop Is vanilla, olives, cocoa, sugar cane, coffee, henequ4n, rubber, coconuts, grapes or fruit trees. There are a number of definitional problems here. it Is not clear If the category "fruit trees" Includes nuts or frults that grow on bushes. Glven that grapes are singled out for special treatment, why not strawberries also? Finally, It Is not apparent whether all forms of sisal fall wlthin the 300 hectare linit, or Just henequdn (Yates,1981a:177). 30. in the case of cattle ranches, the coing Is calculated according to the land area needed to support 500 head of cattle or an equivalent number of small livestock. There is some ambiguity here because, In the 1971 ammendments to the reform law, reference is made not only to the 500-head standard but also to a 20 ceiling of 800 hectares for cattle ranches In areas of poor-quallty pasture. The two speclfications are mutually Inconsistent because It Is impossible to support 500 head of cattle on 800 hectares of the poorest-quallty grazing land. Also, the equivalences between head of cattle and head of sheep, pigs and goats Is nowhere Indicated (Yates,1981a:178). 31. Finally, It Is not clear what Is the maxlnum area of a cattle ranch that may be devoted to crop cultivation without the property being reclassifled as an arable farm (and therefore subject to tighter holding-size llmits). While the legislation originally prohibited stock rearers from growing crops, the 1981 Agricultural Development Law sanctioned "mixed" farming, failing however to specify the acceptable ratio of cropland to grazing land (Sanderson,1984:48). 32. Effective enforcement of these land-use and holding-size restrictions presupposes that the government Is reliably Informed of land-use capabilities throughout the natlon. Responsibility for assessing land productivity rests with the local delegations from the Agrarian Reform Ministry that are based In each state. The survey work that must be carried out before appropriate holding-size limits may be specified remains Incomplete. Therefore, in several parts of the republic, the definitlon of what constitutes an acceptable holding size remains vague. Moreover, actual enforcement of these provisions varles widely between regions. 33. Also, there Is a significant loophole In the legislatlon that explains much of the Inconsistency In the enforcement of holding-size limits. Changes made to the land reform law In 1947 permit large estate owners to obtain legal guarantees (the right of amparo) that their property will not be expropriated for an Indefinite period, pending official Investigatlon of the legitimacy of their hold over land; during this period the landowners are issued with certificates (certificados de Inafectibilidad) protecting their propertles from seizure. This protection Is provided 21 by JudIcial authorities at the state level and has often served as a "preemptive strkoe against the agrarian reform, blocking redistribution before any petitions for land have actually been bdgd. 34. The legislation also contains stipulations about the size of parcels on individual ejidos: according to the Federal Agrarian Reform Law of 1971, parcels should contain ton hectares of Irrigated land (or twenty hectares of rainfed land) (Sanderson,1984:49). This law appears to be honoured In the breach in 1981, the average size of parcels on Individual ejidos was 7 hectares. In most cases, the ejldo land base has proved too small to accommodate all of those petitioning for land In a manner that Is consistent with the rule about parcel size. Indeed, legal provision exists for the land to be parcelled out In smaller units than officially prescribed, when the ejido !s formally Inaugurated (Dlaz Clsneros,1983:21). 35. However, In some cases ejldatarlos have failed to receive formal title to their parcels (titulos parcelarlos) because they are smaller than the minimum size required by the law (Stavenhagen,1975:147). In 1984, the Agrarian Reform Ministry announced that 86 ps,rcent of the holdings In the ejido sector lack secure tenure owing to the absence of certificates of title (Heath,1985:117). In relatlon to credit access, It might be argued that the absence of title to the land Is not a significant obstacle since ejldo land may not, under any circumstances, be used as collateral for securing a loan. However, the possession of title does confer greater security of tenure upon the ejidatarlo, In the context of his status within the Ejldo. The ejido leadership (Comisarlado Ejldal) has the power to Initiate procedures for dispossessing ejldatarlos whose parcels have been left Idle or rented out. These powers are sometimes exercised selectively and arbitrarily. It could be argued that the absence of title makes It more likely that the ejidatarlo may be unfairly evicted If he falls foul of the leadership. 22 (b) Control over the Identity of those working ejido land. Formally, there are binmerable restrictions on the persons eligible to work ejldo land. The legislation Is designed to prevent the transfer of ejido property (or hcome darived from this property) to persons other than the original ejidatarios or their descendants. in this rospect the law Is ful of contradictions, lending Itself to selective intrpretation and enforcement. Article 77 of the Reform Law prohbfts ejkdaos from using hired hands as a substitute for their own labour; however, they may empioy hired workers as a supplement to their own labour Input, providing they themselves are fully employed In working the parcel (Zaragoza & Maclas,1980:175). Sharecropping Is prohlbited, "except In special circumstances" (Cralg,1983:257): for example, If the widow of an ejidatarlo has no sons of working age to work the land on her oehalf. If attempts are made to rent or sell the parcel, or If It Is left Idle for two years or more, the land may revert to the community and be realiccated to other memiers of the ejldo. 36. Of all the formal restrictions on land management, none has generated more confusion than the provisions that ban the renting out of ejido parcels. While all scholars agree that, In practice, renting-out Is widespread (Yates,1981a:181; NACLA,1978:18), there Is a lack of agreement about the circumstances in whIch ttis Is Illegal. Bailey (1981:357), Craig (1983:257) and Yates (1981a:181) flatly assert that renting Is Illogal. However, Bartra (1975:143), Van Ghneken (1980:63) and Zaragoza and Macias (1980:176) indicate that, "in special circumstances", renting may be permitted under the law. 37. Foliowing passage of the 1971 Federal Agrarian Reform Law, the ban on renting was modified slightly: '"hile leasing Is prohibited by Article 55...exceptions are granted In Article 76, which permits parcel leasing and the use of salaried 23 labour. This Is allowed in those cases where the ejidatarlo cannot take on certain tasks profitably, even though he might dedicate his total efforts to them" (Bartra,1975:143). Such provisions give enormous latitude to federal and local authorities In deciding whether specific instances of renting are permissible; since ejidatarios are generally reokctant to admit to renting (Mummert,1987:532) It would appoar that they thomselves are not convinced that the law will back them up. 38. The restrictlon on renting means that ejido land does not necemarily end up In the hands of those with the means and the vocation to work It. Although, In spite of the law, renting Is currently widespread, It Is reasonable to argue that, In the absence of any ban on leasing, the Incidence of this practice would be even greater: there are presumably a number of ejidatarlos (particularly those on bad terms with the ejldo leadership) who are deterred from renting by the prospect that their parcels may be confiscated. Theoretically, restrictions on leasing would lead to subopthnal use of land, driving down the average productivity of ejidos In reoation to private farms. 39. The situation has been further complicated by the Agricultural Development Law (1981) which seeks to promote "production units" In which ejidos and private producers can jointly produce staple crops (Balley,1981:360); the terms of this arrangement were liberalized by legislation In late 1983, making the contract between ejidatarios and private farmers renewable, not annually as before, but once every three years (Heath,1985:118). 40. While the 1981 Agricultural Census sought to determine the number of these "mixed" (ejido/private) production units, the data (Table 4) probably underestkmate the extent to which ejidatarios and private farmers pool resources. This type of associatlon admits of several possibilities. There are probably many cases where the ejidatarlo supplies his land (and possibly his labour) while the 24 private farmer finances capital Inputs. The law envisages that the ejidatarlo shall recelve a share of the profits generated from the association. In practice, who Is to say If he receives an equal share of the profits or Is merely paid the equivalent of a rent for the land? The now legislatlon may be Interpreted as an kupilcit legalzation of the renting out of ejido parcels. 41. On the other hand, the 1981 legislation Increases the powers of the Mkttry of Agriture to seize underutilized and and farm them out to andlee peasants (Bailey,1981:360). The 1981 Census distngulshes, for the first time, between holdings that are being farmed and those that are Idle (Table 5). Thus, ejidatarlos who rent to private farmers and perform no work on the land themselves (preferring to pursue off-farm empioyment) may still run the rlsk of having their land taken away from them. Since there Is no clear-cut distinction between renting out and sharlng in the profits of a Joint venture (both Involve some payment to the ejidatarlo), the new law lends Itself to arbitrary appilcation: ajidatarios have every reason to think twice about collaborating with private farmers, for fear they may have their parcels confiscated. This Is another respect In which the law may adversely affect the average productivity of the ejido sector, by reducing the transfer of land from less efficient to more efficient producers. V. ACCESS TO LAND AND LAND USE PATTERNS. 42. The quality of land distributed under the land reform has deciined since 1940. Before that year, cropland accounted for about 20 percent of all the land handed out; 4 percent of the total land distributed was Irrigated. However, In 1965-76, only 8 percent of the newly incorporated land was suitable for crops, 25 virtually none of It Irrigated. There was some kIprovement after 1976, with 14 percent of the land added to the reform sector being sultable for cultivation (2 percent of all land distrlbuted was Irrigated). Currently, the total cropland In the reform sector accounts for 14 percent of the cumulative land reform area (I.e. taking Into account all land added to the sector from 1900 to 1984XINEGI,1986:273; Table 6). Non-crop land comprises forests, natural pasture, rough grazing and land that Is Inappropriate for any agricultural use.- 43. The 1981 Agricultural Census Indicates that the average size of the ejldo parcel was 7.3 hectares; in the crowded states of Mexico and Puebla (both bordering on the Federal District), average parcel size was respectively 2.1 and 3.5 hectares. While the average holding size for privately-owned land is much larger (73.6 hectares In 1981), Just under 60 percent of both private and ejido enterprises operate small holdings (up to 5 hectares) (Table 7). "Mlni holdings (2 hectares or less) are more common In the private sector (40 percent of all holdings) than In the ejido sector (31 percent), suggesting perhaps that inheritance procedures In the ejldos (the stipulatlon that parcels be passed on Intact to a single heir rather than divided between several heirs) have acted as a partial brake on holding fragmentation. 44. In 1981, proportionately more land was left out of productlon (idle) In the private sector than In the ejldo sector (27 percent and 15 percent respectively)(Table 5). However, If the comparison Is restricted to enterprises no larger than 5 hectares, the proportion of land left Idle was the same for both tenure categories (6 percent). In 5 states, more than half of the land In private holdings was Idle In 1981; In the ejldo sector, the same level of neglect was true only for the state of Sonora (Table 8). 26 45. If these figures are accurate, they depict a massive underuse of agricultural land by larger farm enterprlses. In view of the possibility that Idle land may be expropriated by the state, andowners would surely tend to understate the proportion of tholr holdig not currently In production. Therefore, the census data probably underestkuate the amount of land loft Idle. On the other hand, If the data ar accurate, they still reflect a hlgh level of underexpoitatlon; also, If ndownr did own up to the amount of land left out of production, thth would suggest that they do not treat very seriusly the threat of expropriation. However, In all probabillty, the data are not accurate (see Appendix A). 46. Since the vast majority of ejidos are IndIvIdual rather than collective, there Is some justification for treating the parcel rather than the total land endowment of the community as the basic unit of enterprise. There Is a large gap between the area formed by summing all the parcels In Individual ejidos (equal to 17 percent of land In holdings) and the area covered by the whol ejido sector (ust over half the land In holdings) (Table 8). This gap corresponds prknarlly to the communal lands of Individual ejidos; the collective ejidos contribute to a lesser degree. 47. The considerable amount of land tied up In communal areas raises kmportant policy Issues: If'this land were poorly managed It would greatly reduce the overall productivity of the ejido sector In relation to private farms. From a crop output perspective poor management Is probably not a slgniftcant constraint since much communal land Is steeply sioping with poor soils and and lknited crop potential: even If management were significantly kmproved the kmpact on crop output would probably be lklited. 48. However, the returns resulting from hmproved management of communal forest and rough grazing land would probably be considerable. At the moment, 27 forests are mined Indiscriminately both for commercial purposes and for domestic heating and cooking fuel. The potential for controlled, ecologically-responsbie tilior extraction Is underexploited. One of the prkne purposes of the Agrcuitural Devoepment Law (1981) was to use joint ventures between ejidos and the private sector to Increase the efficiency of forestry activitlis: "although three- quarters of forest property Is In the hands of ejldos they contributo only 17 percent to the output of forest products; In contrast, private enterprise owns 20 percont of forest property and accounts for 65 percent of forest production (Excdisior,19&20 March,1984). 49. Resource depletion from overgrazing Is also a serious problem. Yates Indicates that In 1970 the available grassland per "herblvorous anknal unit" was roughly half as much in the ejido sector as in the private sector (1981a:94). However, Yates Is referring to the private and ejido sectors as a whole: If the comparison were restricted to small farms In each sector there was probably less overgrazing In the ejidos (because, unlike small private farmers, ejidatarios have access to communal grazing land). 50. Between 1970 and 1981, the cattle herd In the ejido sector Increased by 61 percent (compared to only 8 percent In the private dector); the area per head of cattle (total area, not grassland) declined from 17.3 to 13.3 hectares In the ejldo sector. The ejidatarlo's reliance on communal pastures Is reflected in Table 9 which shows the proportlon of the holding In different types of land use: only 14 percent of the ejido parcel Is occupied by natural pasture compared to 37 percent In the case of private holdings. 51. Both In terms of total herdsize and the number of purebred cattle the private farms control more head than the ejido sector. However, on farms up to five hectares the mean number of cattle per enterprise was roughly the same for 28 both ejido and private enterprises In 1981 (Table 10). This means that, focussing on comparably sized farms, there appears to be no signlficant difference between tenure categories in wealth terms. 52. There Is no significant difference in the proportion of cropland under fallow between tenure categories (16 percent and 17 percent respectively, for private and ejido enterprises). However, a much larger proportion of the land in private enterprises has no defined use (13 percent compared to 3 percent for ejidos). suggesting a lower intensity of agricultural exploitation. If the comparison Is limited to small farms (not exceeding fIve hectares), the ejidos also have a lower proportion of land with no defined use, but the gap between tenure categories Is less significant (4 percent and 1 percent undefined, for private farms and ejidos respectively) (Table 9). 53. Only to a limited extent may this difference In the level of exploitatlon be accounted for by variations In average land quality between the two tenure categories. Access to Irrigation no longer varies significantly between tenure categories (Table 11). In the case of enterprlses up to five hectares, elido parcels are slightly better served than private parcels (respectively 18 percent and 13 percent were irrigated). This Is a significant change on 1970 when private enterprises up to five hectares had more priveleged access to Irrigation: 24 percent of arable land Irrigated compared to 15 percent In the case of the ejido sector. 54. There are considerable variations between regions In the proportion of cropland under Irrigation (Table 12). In the Gulf and South Pacific reglons where low and erratic rainfall Is less of a constraint, under five percent of ejido cropland Is Irrigated. In the North Pacific region, where the climate Is so arid that Irrigatlon Is Indispensable, the private sector Is much better served by irrigation 29 than the ejido sector: respectively 60 percent and 46 percent was irrigated in 1981. Betwoon 1970 and 1981 the proportion of cropland under Irrigation actually declhed (for both ejidos and private farms) in the North Pacific and North regions. hI the Contre, the proportion of land Irrigated grow in the case of ejidos, but shrank In the case of private farms. This marks an kmportant reversal of pro- 1970 trends (Wionczek,1982). 55. At a natikoie level, there is not a silgnificant difference In the croppng patterns of the private and ojido sectors: in both cases, the three crops that cover the largest area are respectively maize, beans and sorghum (Table 13). In 1981, these crops accounted for 75 percent of the area harvested In the case of private holdings and 84 percent In the case of the ejido sector. The cropping pattern Is less diversified on ejidos and, In the aggregate, generates lower revenue per hectare than the private farms. The lower level of diversification may partly reflect the state credit bank's control over the cropping pattern: producers working with Banrural (three-quarters of whom are ejldatarlos) have less freedom of choice over the crops they cultivate. VI. LABOUR ORGANIZATION AND EMPLOYMENT 58. In general, there Is little or no difference In mployment patterns between Wxdlvidual ejidos and villages consisting of small private landowners: the ejido leadership has no Influence on the amount of work that households perform on their individual parcels and the amount of work performed by collective labour gangs, at the behest of the Comisarlado, (e.g. maintenance of communal Infrastructure) tends to be limited. 30 57. According to the 1981 census, In individual ejidos, 50 percent of jildatarlos employ casual wage labour at some tkne In the crop cycle. However, these data shed no light on the number ot days per year durkn which labour is hired In (hiring-in Is usually restricted to seasonal peaks, particularly harvesting). The same source Indicates that only 6 percent of ajidatarlos are Involved In cooperative forms of production with other households (Table 15). This Is undoubtedly a gross underestimate and may reflect a tendency for ojkdataros to conceal practices whose legality Is In doubt. The data almost certainly fall to reflect the frequency of sharecropping arrangements among ejldatarlos, a practice that case-study evidence Indicates to be widespread (Flnkler,1978; Gregory,1986; Mummort,1 987). 58. Finally, the available labour use data are of Iklited use because they refer only to labour employed on the farm: they provide no way of assessing the role of off-farm employment In the total labour profile of the ejido household. 59. In all ejldos, the employment pattern Is highly diversified. Diversity Is a response to three circumstances. First, there Is considerable wealth differentiation within the ejldo, Initially a consequence of the differences In parcel size and land quality: latecomers to the ejido are generally worst off In terms of resource endowment. 60. Second, the majority of ejldatarlos In Mexico control too little land to depend entirely on this as a source of Income: they are obliged to look for off- farm work, either as casual agricultural labourers or In the urban sector (particularly In construction). In one study, conducted In the 1960s, It was calculated that 84 percent of all ejido plots were too small to provide full employment and an adequate Income for a peasant family; the proportion was roughly the same (85 percent) among private farms (Stavenhagen,1975:146). 31 61. Recourse to off-ftam work has been further encouraged by urban real wage trends which (until recently at least) have been more favourable than trends in the farm-gate prkces of crop staples (Graph 1). Moreover, since only 14 prcent of the land in the reform sector Is suitable for cultivation (INEGI,1986:273), the agricultural employment capacity of the ejido sector Is severely constrainod. 62. Third, there Is an inbuilt employment problem In the ejido sector because only one child Is permtted to inherit the family parcel: the dispossessed younger sons and daughters have to seek alternative employment. Some of this surplus labour has been absorbed by agroindustrial activities conducted at the ejldo level; agroindustries (e.g. cotton ginning or henequ4n defibratlon) are most developed on the collective ejidos (Wllkle,1971). The fierce competition for jobs has often resulted In overmanning and the appearance of "phantom workers" on the payroll of collective ejidos (Brannon & Baklanoff,1984). Generally, however, on both indIvidual and collective ejidos, children falling to Inherit a usufruct title must seek work outside the ejido sector. 63. This Is kmportant because It means that most ejldo households have one or more members engaged In off-farm work, frequently Involving temporary (often seasonal) migration. Migration leads to the develpment of a family network that links even the most remote rural areas to metropolitan centres in Mexico and to the U.S. labour market. The development of such a network makes migratlon progressively easier by providing new migrants with places to stay, emotonal support and informatlon about Jobs. Rural outmigratlon Is offset to some degree by return migration; moreover, those who move permanently to the towns often contrive, via the urban-rural kin network, to retaln a toehold on the land (Arizpe,1 981 ). 32 64. In his study of an ejido in Mexico state, De Walt (1979:220-222) found that 85 percent of the ejidatarlos had worked In neighbouring Mexico Clty at least once, usually on a seasonal basis; moreover, 72 percent of ejido members had at least one child living In the capital city. Both De Walt (1979:221) and Roberts (1982:310)-the latter writing on ejidos In the BaJio region-report the absence of a signifcant correlatlon between mlgration and wealth: specifically, the propensity of ejidatarios to engage in off-farm work does not vary inversely with farm Income. 65. The kIportance that farmers attach to off-farm works helps to explain why farms are tending to employ less labour per hectare. In the 19709 there was a significant expansion In the number of enterprises using tractors; the growth rate was higher In the ejido sector than the private sector, to compensate for the relative scarcity of tractors In the elido sector before 1970 (G6mez Oliver,1984; SPP-UNIDO,1986). Between 1970 and 1981, the share of cattle used as work ankials halved for both the ejido and private sectors (Table 18), reflecting the substitution of tractors for ploughing, cultivation and carting. 66. In the same period, there was an expansion of livestock activities and a move toward less labour intensive crops (notably sorghum), trends that were as marked for small as large farmers (CIMMYT,1974:7; Norton,1987). Government programmes whose purpose was to promote small-farmer adoptlon of higher-yielding maize varieties encountered some resistance to the Introduction of new technologles Involving higher labour Inputs per hectare: It has been argued that this resistance reflects farmers' calculations that they may earn more In off-farm than In on-farm work (Villa Issa,1977:216; Redciift,1983). 67. The tendency for farmwork to employ less labour per hectare means that there Is reduced opportunity for ejldatarios to supplement thelr farm Incomes by 33 working as casual day labourers for other farmers: there has been a decline In th employment of hired hands In agriculture (Gregory,1986). A recent study of an eldo In Michoacin demonstrated that only 9 percent of farmers hired in labour: increased use of tractors had reduced the need to hire In labour; also, households with insufficient family labour to carry out farm work usually prefetrred to enter Into sharerop arrangaments with "labour-surphks households, rather than eploy wage workors (Mummrt,1987). Comparing this study with a t956 study of the same ejido (Frledrlch,1970). It Is clear that there has been a decline in the number of ejidatarlos mploying temporary wage labourers and a rise In sharecropping (Mummert,1987:536). 68. In the cash-poor, high-risk environment of most ejidos sharecropping makes better sense for labour-deficit households than hiring In wage workers. Typically, sharecropping entails households that are short of labour handing over land to "labour-surplus' households wlth the latter assuming full responsIbility for cultivatlon. The two households share the harvest (and the risk of crop fallure) equally between them. The household supplying land Is usually responsible for supplying draught anknals, seed and, In some cases, fertilizer. In most cases, the collaborating households belong to the same kinship group (Mummert,1987:529). 69. The labour-surplus, land-deficit households that are actively looking for land to work are not necessarily the poorer or less powerful party to the arrangement. A study of Irrigated ejidos In the Mezqultal Valley (Hidalgo) reveals that these households may be highly entrepreneurial, simultaneously Involved In a number of sharecropping deals. Thus, "sharecropping provides an avenue for economic kIprovement" (Flnkler,1978:105). 70. Sharecropping enhances the labour mobillty of ejidatarlos: It enables them to share In the output of their land while freeing them to pursue off-farm work. 34 The same effect Is achieved by the leasing-out of ejido parcels, a practice which appears to be at least as widespread as sharecropping. In Mummort's study, only 3 percont of households admitted to renting out their and but, according to key Informants In the ejido, no iess than 30 percent of ejidatarlos were actually engaged in loashg-out. In this ejido, renting-out Is percelved as a last resort because the anmual rent received amounts to less than the expected valu of the maize crop that wiN be cultivated on the rented land (Mwuert,1987:532). 71. It is mainly the poor who engage In renting-out because they cannot afford to sharecrop: In other words, the off-farm work on which they depend for survival generates the Income to cover family subsistence but Is not sufficient to cover the Input purchase entailed by sharecropping. While sharecrop arrangements take place primarily between different households In the same ejido, rental arrangements cover the leasing-out of ejido land to private farmers as well as transactions between ojldatarlos. 72. Although It Is kIpossible to quantify the extent of renting-out, there Is a broad concensus that the practice Is so widespread as to open up an enormous gap between the formal distributlon of land between tenure sectors (as reflected In the census data) and the real control that ejidatarlos and private farmers exercise over resources (Reyes Osorlo etal.,1974:451; Zaragoza & Maclas,1980:171; Warman,1977). One source esthuates that the llegal renting of ojIdo parcels affects ufrom 50 to 90 percent of all ejido land In several Irrlgation districts In Sonora and Sinaloa, and from 35 to 50 percent In the BajIo, In Mlchoactn and Jalisco" (Yates,1981a:181). The North American Congress for Latin America reported that, "In Sonora, about 70 percent of the ejidos are rented, especially In the irrlgated areas of NavaJoa and the Valle del Yaqul, while esthiates for... Sinaloa range between 40 and 80 percent" (NACLA,1976:18). 35 73. To concluds, the labour-substituting trends In agriulture, coupled with the frequency of leasing and sharecropping, argue strongly that the land reform has not served to *tbe the rural population to the land, but has bon associated with an accelerating process of rural outmigratlon. There Is no Indication that the intensity of this process varies signiflcantly between ejidos and small private farms. The Incidence of outmigratlon Is probably the same for both tenure categorles sico the mean holding size Is the same, the prevalence of minfundla restricting the scope for on-farm employment expansion. On the other hand, while average land quality Is now roughly the same for ejido and small private farm sectors, the proportion of land left out of agricultural use Is somewhat higher In the case of small private farms (according to the 1981 census data). VII. ACCESS TO CREDIT 74. Compared to small private farmers, ojldatarios depend more on state- lending Institutlons for credit because the agrarian reform law prevents tham from mortgaging their land: since they cannot use their parcels as collateral It Is difficult for them to secure loans from commercial banks. The agricultural development bank, Banrural (in common with previous state Institutions that lent money to the ojidos) waives the collateral requirement and exercises weak sanctions against loan default. This Is one aspect of the paternalistic relatonship between state banks and ejidatarlos, a paternalism that also colours Banrural loans to private farmers. 75. The agricultural development bank, Banrural, accounts for about 45 percent of total lending to agriculture and Is the main source of credit for small farmers In both the ejido and the private sectors: about three-quarters of Its 36 clients are ejldatarios and the credit Banrural extends to this group amounts to roughly one-third of all loans to agriculture (Table 17). 76. The paternalism Inherent In development bank lending reflects an ethos that has Its roots In the Mexican land reform: polltical considerations tend to assume precedence over economic ones; credit, like land, Is used to buy the political support of the peasantry rather than primarily to enhance agricultural productivity. This approach may be traced back to the 1926 Act which first established an Institutkonal credit facility for the ejidos. According to Reolo (1987:30), the farmer groups that emerged at this time as the prkmary channels for state credit were essentlally created by the government and served to enhance its powerbase In the countryside. The collective ejidos that were established by C4rdenas In the 1930s developed out of these state-implemented credit associations (Hewitt de Alcintara,1978:69). 77. The government operated on the assumption that peasants were likely to remain, first, poor and, second, averse to adopting new technology: therefore, the state would likely remain their only source of loans and It would also have to assume responsibility for providing the appropriate Inputs. Little incentive was given to peasant farmers to mobilize savings on their own behalf, nor little credence given to their ability as farm managers. This attitude affected the state's deallngs with both small private farmers and ejldatarlos and Is alnost as widespread today as It was In the early period of the regime. 78. The negative Impact of this paternalistic approach Is currently discernible, first, In decislons about the crop mix and, second, In arrangements for Input provision. Banrural exercises considerable Influence over the cropping pattern since only certain crops (basically, those In the guarantee-price regime) are eligible for financing. The crop pattern tends to be less responsive to price 37 trends than It would be If the bank allowed farmers to decide for themselves what crops to finance with the credit resources they receive. The adverse consequencos of this pollcy are evident In the collective ejidos (which are essentially managed by Banrural): In La Laguna and the Yucathn these ejidos conthue to monocrop (cotton and henequen respectively), despite a progressive deterioration In the relative profitability of these crops (Wlikle,1971; Brannon & Baklanoff,1984 L 1987). 79. Even on the Individual ejidos and among small private farmers working with Banrural, the cropping pattern is unnecessarily rigid, In the state of Morelos, for example, the state's credit policies (reinforced by cropping restrictlons defined by the land reform law) obliged ejldatarlos to continue growing rice and sugar, despite the declining profitability of these crops. Private farms In the same regions (working with commercial banks, not Banrural) showed greater Inclination to diversify (Strlnger,1972). 80. Second, Inputs sUch as seed, fertilizer and pesticide are provided In kind by Banrural, on the assumption that credit resources are more likely to be misappropriated If farmers are given cash with which to purchase their own Inputs. However, there Is abundant evidence to Indicate that farmers often sell the Inputs they receive from Banrural, exchanging them for cash on secondary markets (Relio,1987:65). A further problem arises from Banrural's failure to ensure tknely delivery of Inputs: famers frequently complain that fertilizer and pesticide arrive too late In the crop cycle to be fully effective. 81. Also, partty In order to achieve scale economles In procurement, the Bank operates with a standardized Input package that Is Insensitive to regional variations In Input requirements and prices. These variations are significant at a micro-regional as well as a regional level: no opportunity Is given to farmers to 38 boost yields by carefully tailoring the Input package to the precise needs of their land; no confidence Is placed In their ability to Identify these needs. For example, studies indlcate that farmers have little confidence In the sorghum varieties supplbed to Banrural by the state-owned seed manufacturer, PRONASE; they often prefer to buy from commercial manufacturers, finding that these seeds offer higher yields (Sulrez,1982-83; Rello,1987:85). 82. Although in these respects, ejkdatarios and private farmers lendng from Banrural face the same constraints, there are two obstacles to credit access that apply specifically to ejidatarlos. First, unlike private farmers seeking working- capital loans from Banrural, ejidatarlos cannot make Individual credit applications: credit Is negotiated collectively through the ejldo leadership. Individuals seeking credit need the backing of the leadership and If, for personal or polltical reasons, they have a confilct with the Comisarlado they may be denied access to credit. By the same token, the political allies or favourites of the ejldo leadership may obtain preferential treatment, irrespective of thelr creditworthiness or their past repayment record (Relio,1987:84). 83. There Is a long tradition of ejido leaders being coopted by the banks. This contributes to Infighting within the ejido, causing many ejidatarios to feel little Identification with the ejldo as a whole and making them disinclined to adopt a cooperative or responsible attitude toward the bank. In many ejidos, breakaway factions have formed: disgruntled with the corruption of entrenched leaders who are In cahoots with bank officlals, ejldatarios have sought to establish independent credit unions. The formatbon of such Independent unions has often met with brutal repression because they threaten the power base of local leaders (Wikie,1970). 84. Second, If some ejidatarios fall to repay their loans at the end of the crop cycle, the bank may respond by refusing to extend fresh credit to the ejido 39 as a whole: thus, the actions of an Insolvent minority may harm the interests of more creditworthy individuals with a good repayment record (Rello,1987:84). The bank may feel that defaulters are more likely to pay up If faced with pressure from angry peers who have been cut off from credit. However, experience in various countries demonstrates that effective pressure Is only brought to bear when the group Is small and tightly knit (Adams & Vogel,1986). 85. in this respoet, the ejldo Is often an Inappropriate unit for credit alocation purposes. The members of an ejido do not in most cases work together on a cooperative basis, either for production or marketing purposes. The frequent renting out of parcels to outsiders and the high absenteelsm of ejidatarlos (owing to temporary migration) help to undermine solidarity between ejido members. Moreover, group solidarity Is likely to be diluted by the sheer size of most ejido coumunitles and the cilentelist policies often pursued by the ejido leadership. 88. This helps to explain the lack of credit discipline in state-bank lending to 3Jidatars. The only sanction that Banrural may bring to bear on ejidatarios who fal to repay their ioans is to deny them further credit: they cannot seize their assets and In many cases there Is no crop to seize either. This Increases the likelihood of Individual default. Glven that these Individual cases of default make It likely that the whole ejido will be cut of f from credit In the next crop cycle, there Is ittle incentive for the solvent to repay their ioans. A recent survey of rural financial markets concludes that "the most kmportant factors In determining whether a loan was likely to be delinquent were those related to the borrower's assessment of the probability of obtaining a new, larger loan In the future on a thiely basis" (Adams & Vogel,1986:483); even the most creditworthy ejidatario is likely to have low expectations in this respect. 40 87. The likelihood of individual default Is further Wicreased by the tendency of Banrurai to spread credit resources too thinly: rather than restricting crodit to producers with productive potential Banrural Is obliged, for political and social welfare reasons, to operate with a large number of poor farmers. Partly because It continues to finance a large number of marginal producers, Banrural has responded to post-1981 resource constraints by reducing Its long-term inveent lendhg to a much greater extent than Its short-term working capital loans: the average volume of Investment loans tell by 58 percent between 1980- 82 and 1986-88, compared to 20 percent In the case of working-capital loans (Table 18). In the 1980s, despite severe funding constraints, there has been an Increase In the proportion of land planted to maize that Is covered by Banrural credit: from an average of 34 percent of the sown area In 1980-82 to 39 percent In 1986-88 (Banrural,1989). Since Mexico's marginal farmers are primarily maize growers these data suggest the extent to which Banrural loans continue to t e directed toward poorer producers (mainly ejidatarios). In the case of ejidatarios (unlike small private farmers), loan provision Is not based on Individual creditworthiness, Increasing the likelihood of default. 88. State lending policy reflects a confusion between the objective of poverty alleviation (using short-term credit as a transfer payment) and the objective of enhancing agricultural productivity (Goodman et al,1985:71). Hewitt de Alc&ntara (1978:87-89) observes that the credit policy of the development banks has fluctuated between the objective of providing for broad-based communal development and the objective of restricting credit to those with the means to repay. When the first of these objectives predominates (e.g. In 1980-82), there Is an enormous burden placed on the financial health of the bank as many loans are not recovered. Currently, the supply of fresh credit Is severely restricted by 41 the funds that have to be set aside to cover overdue loans and to finance rescheduling. The effect of spreading credit too thinly Is to lkhit the productivity of agricultural credit. While the problem arises with small farmers I general, It is partcutay acute In relation to ejidatarlos. Vill. THE EJIDO MARKETING SUPPORT PROGRAM 89. Partly in response to rural social upheaval In the early 1970s, the govermnent launched the Ejldo Marketing Support Programme (PACE) In 1975 (Fox,1986:197). Although originally targetted at ejidatarios, the programme was subsequently expanded to all small farmers. Therefore, the problems described here are not a consequence of institutional blasses associated with the land reform: they apply equally to ejidatarios and private farmers. 90. PACE was conceived as a means of bolstering the Incomes of small-grain producers by reducing their dependence on Texpioitative" middlemen operating at the village level (Boruconsa,1982:16). By facilitating access to rural warehouses owned or leased by the state (under the aegis of Conasupo), PACE was intended to provide alternative marketing channels to ejidatarlos, purportedly makIng It harder for middlemen to force low prices on producers and thus increasing the effectiveness of the state's guarantee price policy. The ultinate objective was to sthnuWate output growth by making It easier for small producers to generate savings from staple production, enhancing prospects for on-farm Investment (Fox.1 986:199). 91. PACE focusses on basic grains and akns to relmburse producers for transport and handling costs Incurred in delivering their grain to state warehouses. The level of freight reinbursement Is based on the distance between warehouse 42 and field and weighted according to the quality of road access. Transportation costs tend to be significantly higher on ungraded roads, not only because of accelerated vehicle deprociatlon and ongor journey thne, but also because on these routes there Is much less competition betwoen freight handlers (FoX,1 986:204). 92. in order to dicourage wholesalers from taking advantage of this markerhg subsidy, robdbrsment Is, In prOcipl, restricted to those delvorhn ess than S0 tonnes of grain pr year to the rural warehouses. Conasupo lends sacks to farmers to facilitate transport and handling. The rural warehouses are located within or close to an ejido and staffed by ejidatarios nominated by the members of the ejido; Conasupo trains the warehouse hands In book-keeping, quality control and fumigation procedures. 93. The kIpact of this programme Is partly contlgent upon the extension of the rural warehouse network. Warehouses have, for sound economic reasons, been located primarily In areas where the largest grain surpluses are produced. They tend to be few and far between In poorer, more remote regions; yet, In terms of PACE objectives, It Is In the poorest areas that access to warehouses needs to be facilltated, precisely because It Is here that the middlemen have the strongest monopoly over marketing and producer Incomes are most In need of enhancement. This contradiction has never been resolved. 94. The poor promotion of PACE means that It covers only a minority of producers. Moreover, many of those registered In the programme do not actually participate In It, preferring to sell their grain to private traders. Less than 10 percent of those who signed up for the programme In 1975-76 ultbnately sold grain to Conasupo (Fox,1986:200). In 1984-85, only 16 percent of the MIchoac&n producers registered In PACE actively participated In the programme 43 (Heath,1987:278). Admittedly, these figures are not a perfect measure of the programme's kmpact: easier access to Conasupo warehouses may enable producers to lever middlemen Into paying them an acceptable prico, a benefit enjoyed both by participants and non-partIcipants In PACE. However, a further reason for low participation rates Is that, In the absence of thuely credit facilitles, some producers are driven to seek loans from middlemen, committing their harvest In advance (invariably at a prico wel below the official lovel). 95. A study In Zittcuaro, Michoacgn, demonstrates that, even In areas with a good warehouse network and excellent road access, most ejidatarios are disinclined to sell their grain to the state. The quality controls operated by Conasupo entail payment deductions for grain that Is dirty, damp or pest-ridden. Ejldatarios argue that private traders Impose less domanding product standards. A further objection concerns the great expenditure of time Involved In selilng to Conasupo (there are long queues at the warehouses In the kmmodlate aftermath of the harvest) and producers' understandable preference for on-the-spot payment In cash. 96. By and large, In this area the programme seemed to cater for producers with more than two tonnes to sell. The author concludes that only farmers with more than two tonnes to market (a minority of the farmors In this area) do fInd It worthwhile to sell to Conasupo. The low level of direct sales from farmer to state Is loss a reflection of problematIc access to the warehouses than It Is a consequence of the llmited number of farmers producing a significant marketable surplus of maize. This In turn Is attrlbutable to low yields (exacerbated by Ineffective credit policies) and Inadequate price incentives: these factors appear to pose more fundamental constraints than the nature of marketing facilities (Heath,1 987). 44 97. in the rich agricultural state of Chihuahua, Fox (1986:208-209) discovered that middlemen evade the 50-tonne ceiling on grain sales by making a serls of relatively small deliveries to separate but close-together warehouses, solling under the names of clients and family members. To conclude, the marketing subsidy operated by PACE manifests the same contradlctions as subsidized credit programmes: It was Initially designed to benefit the poorest producers but has ended up providing an unnecessary cushion to commercial-scale producers and wholesalers, particularly In those more prosperous agricultural regions where the warehouse network is most fully developed. IX. COMPARING THE PERFORMANCE OF EJIDOS AND PRIVATE FARMS. 98. From the discussion In the preceding sectlons It Is possible to frame a number of hypotheses: 1. Ejldatarlos have less security of tenure than private farmers because they often lack usufruct title to their land, making It more likely that they may be arbitrarily evicted; In the event of eviction they would receive no compensation for improvements made to their holding, thus reducing the incentive for on-farm investment. 2. Unclear laws conceming the circumstances In which ejldatarlos may legitinately rent-out their land or engage In sharecropping Inhibit the transfer of land to the most officient producers; accordingly, there Is no process of enterprise concentration In the ejido sector, falling to offset the adverse impact of holding fragmentation on production. 45 3. Mismanagement of communal land In the ejldos depresses the average productivity of this sector In relation to private farms. 4. Ejidatarlos face greater difficulties in access to credit than private famers: first, because they cannot use their land as collateral, cutting them off from commercial bank loans; second, because lending by Banrural has to be arranged thorough Ejldo leadership. This Introduces a possibly non-economic element in the docision rogarding credit alocation. Hypotheses (1)-(4) each suggest that land productivity will be lower on average In the ejido sector than In the private farm sector. There Is one countervailing hypothesis: S. Stipulations in the land reform law concerning holding-size limits and land use restrictions are complex and contradictory: this makes it difficult for private farmers to Judge when they are contravening the law and when they are likely to face expropriation; this will reduce the Incentive for large private farmers to Invest In their land, pulling down the average productivity of the private farm sector In relation to the ejido sector. 99. There are no recent studies of the relative productivity of the two tenure catgories: the avallable sources are based either on agricultural census data up to 1970 or pro-1970 case studies. (Appendix A explains why data from the 1981 Agricuitural Census does not lend Itself to productivity comparisons). To assess the findings of the avallablo studles, It Is necessary to understand the policy onvironment that prevalled before 1970. 100. In the immediate aftermath of the dynamic phase of the Mexican land reform (1934-40), Mexico experienced impressive agricultural growth. Many authors 46 (e.g. Flores,1959 & 1969) have argued that land reform, by ensuring more Intensive cuWtivatlon of large tracts that wore formerly Idle, was the key factor In this growth. Between 1940 and 1960, Meoxcan agriultural output rose at an annual average rate of 4.6 percent (compared to 2.7 percent for al Latin AmericaXEckstohn et al,1978:36). Roughly half of this growth was attributable to an extenion of the cultivated area, the other half to hcroased yields (Reyes Osorbo et al,1974:112). 101. The nature of the land reform process helped to maxknise the contribution that expropriated estateholders made to 1940-80 agricultural growth. "The policy of leaving owners with generous reserves guaranteed against further expropriation transformed, on a significant national scale, a class of owners of very large traditionalist latifundla, Into a group suddenly reduced In their holdings but still able to muster substantial resources and recover their economic position by huprovement In their farming methods. The shock of expropriation would surely provide a now incentive for them to attempt to do this; and the growth In output resulted In large measure from Just this kind of Innovation by larger landholders" (Eckstein et al1978:39). In conclusion, "the landowners' reserves policy probably raised agricultural output In the longer run; while the creation of small ejido family farms increased food production and ejido Incomes In the short run, but kept them at a very modest rate of growth thereafter" (Eckstein et al,1978:42). 102. An important factor tending to enhance the different growth trajectories of ejidos and private farms was the scale-blas that government pollcy imposed on the delivery of subsidized capital Inputs: this bias favoured the large private farms and the collective ejidos, reflecting a government convictlon that large units were Intrinsically more efficient. Johnston and Kilby (1975) have referred to the dualistic or "bbnodal" structure that government policy Induced In MexIcan agriculture: large 47 commercial farms (centred on the Irrigation districts) generated rapid agricultural growth; these coexisted with a subsistence sector of low growth and Increasing poverty, consistng of small farms in the ejido and private sectors (contred on the rain-fed agricultural regions). 103. Agrkiutural credit subsidies, public Investment (notably In irrgation and roads) and gre.n revokution technologies were prkwarily channeled toward larger enterprbse (Hewitt de Alckntara,1 978). In the 1940s and 1950s, the ejldo sector received respectively only 21 percent and 22 percent of total agricultural credit (MueSler,1970:258-257); yet Its share In total arable land was 44 percent In 1950 (Yates,1981a:154). In the same decades the total area of ejido land that received fertilizer was only half that In the private farm sector (Mueller,1970:258). However, smail prtvate farms were not much better off than ejidos. 104. To conclude, up until 1970 at least, private farms over five hectares remained more capitalized than the ejidos (both In terms of working capital and fixed capital stock) (Hewitt de Alcant&ra,1978). The large private farms and (to a lesser extent, the collective ejidos) contributed a disproportionately large share of farm production and agricultural output growth up to 1970 (Eckstein et al,1978:38). (a) The census record. These trends would suggest that any study comparing the performance of ejidos and private farms would surely point to the superior productivity of the latter. However, the published works referring to the pre-1970 perlod reach varying conclusions. Comparisons between ejldos and private farms that are based on the census data have focussed on differences In (a) crop yields, (b) labour productivity, (c) the relationship between purchased capital Inputs and output value and (d) the relationship between an aggregate measure of land, 48 labour and capital, and output value. On each of these counts there Is not complete agreement between the various studies concernIng the productivity of the two tenure sectors. 105. Which Is the most logithnate (or reliable) Indicator of relative productivity? Yates (1981b:748-750) argues that, since the relative scarcity of land Is now probably the same for both the ejido and the private farm sectors, It Is logithate to focus on land productivity when comparing the performance of the two sectors. Moreover, the most recent ovidence suggests that, comparing same-size holdings, there Is no significant varlatIon between tenure categorles In average land quality (Tables 9 & 10). However, It has been suggested that, In the pre-1970 period, Institutional blasses tended to starve the Individual ejldos of good-quality (Irrigated) land (Reyes Osorlo et al,1974; Hewitt de Alcintara,1978); this needs to be borne In mind when assessing the pre-1970 evidence on relative productivity. 106. Dovring (1970) found that the rates of growth of crop output were about the same for both sectors In the 1950s; referring to the same decade, Mueller (1970) made a sknilar fInding, concluding that, In terms of output growth per unit of total measured Inputs, ejidos performed at least as well as private farms. 107. There Is less agreement about relative crop yields. Dovring (1970) concludes that, In 1960, there was no significant difference In crop yields between ejidos, on the one hand, and private farms over five hectares on the other. Also referrlng to the 1960 census, Ecksteln and his associates find that, In terms of physical crop yields, ejldos are less productive than either small or large private farms. However, In terms of the per hectare value of crop output, ejldos are superior to small private farms and Inferior to large private farms, reflecting the proportion of high revenue crops In the crop mix (1978:Appendix C,p.1). 49 108. Analysing the same 1960 census data, Reyes Osorbo and assoclates (1974) derive opposite concluslons: an examination of yield levels by state, controlling for differences In the access to Irrigation between tenure sector, reveals that, except In the case of wheat, ejidos have higher land productivity than large private farms. A study by Pillet (1973) found that, according to the 1960 census, crop yields were roughly the same In both tenure sectors; moreover, the hcrease In yields between the 1950 and 1960 census was greater for the ejidos than for large private farms, In spite of the latter's higher consumption of purchased Inputs. 109. Studies comparing the 1960 and 1970 censuses reach equally varied conclusions. Nguyen and Martinez (1979) find that, In the 1960s, the growth rate of output was considerably greater for the land reform sector than for the private sector; they attribute this, In large part, to the substantial transfer of land from the private to the ejido sector In the 1960s. These authors find that ejidos had lower crop yields than private farms In 1960 and 1970. However, both average and marginal returns to expenditure on farm Inputs were found to be considerably higher for ejidos than for private farms. Referring to the 1970 census data for the state of Puebla, Soberon-Ferrer (1986:165) reaches a different conclusion: he shows that small prIvate farms obtained the highest average and marginal product from land, followed In descending order' by large private farms and ejidos. 110. In terms of the efficlency of resource use, studies based on the census data up to 1970 tend to agree that small farms (ejido and private) are more efficient than large farms. Estimates of the relative efficiency of small private farms and Individual ejidos vary considerably between studies. Eckstein and others (1978:Appendix C,p.1), Wecksteln (1970:405) and Van Ginneken (1980:189-192) find 50 that, hI tms of output per unit of purchased Input, the smai private farms were more efficient than the ekdos. On the other hand, NgUyen and Martinez (1979:831) find that, according to the 1960 and 1970 data, the ejldo sctor obtained a higher crop output value pr unit of capltal expendture than private farms; howver, they compare ejldos with private farm over five hocatares which bi arguably not the most rigorous test of relativo effiency. 111. Efficiency estimates place the ejido In more favourable light If the yardstick Is restricted to purchased capital nputs, If relative efficiency Is measured hI terms of total factor productivity (land and labour as well as capital), the ejido performed less well than the small private farm in the pre-1970 period. in particular, relative efficiency estkiates are highly sonsitive to the value Imputed to unromunerated family labour: If these Inputs are assigned a positive opportunity cost (a reasonable assumptlon In the light of Gregory's (1986) study on the Mexican labour market), then ejidos emerge as much lss officlert users of resources (Nguyen & Martinez,1979:631; Wockstehi,1970:405). Eckstein and assoclates conclude "when total factor productivitles are compared, the small private farms come out ahead in all cases. The relative position of ejidos and large private farms varies with the Inclusion or exclusion of owners' family labour as an hput cost" (1978:115). 112. These conclusions are based on the assumption that the census data are reilabie and convey an accurate picture of trends hI the Mexican countryside. Yates (1 981 a;1981b) provides an Interesting discussion of the problems associated with the census data. He Indicates one respect In which the census data on outputs and Inputs are likely to be Inaccurate: "The private farmers have a strong motive (income tax) for exaggerated reporting of expenses, just as they have for under-reporting receipts, whereas ejidatarios, enjoying statutory exemption from 51 Income tax, have no such motive. Also, In Irrigation districts, Inputs on land leased from ejidatarios are recorded as Incurred by private farmers while the outputs, as we have seen, are credited to the ejidatarlobs (Yates,1981b:747). 113. These points are not entirely convincing. First, Independently of the tax- evasion Issue, ejidatarlos may well be as evasive as private farmers In answering questions about their Income, put to them by persons clearly associated with the government: In the case of rental or sharecrop deals between ejidatarlos, both parties may choose to understate the output from land Involved in these deals (because the deals are Illegal), leading to an understatement of total ejldo output; the evidence (e.g. Mummert,1987) suggests that deals between ejldatarlos may well be as widespread as deals between ejidatarios and private farmers. 114. Second, In cases where ejldatarlos rent-out their land to private farmers, Is It always true that "outputs ...are credited to ejidatarlos"?; Yates reasoning may be neat but reality Is often more mossy. However, there Is one particularly significant point that may be derived from his discussion: In the context of widespread clandestine leasing of ejido parcels (coupled with sharecropping), the formal (de Jura) definition of the size of each tenure sector bears little relation to the de facto control over resources exercised by ejidatarios and private farmers. in other words, apart from being outdated (failing to shed any light on trends since 1970), the census data provide no firm basis for reaching conclusons about the relative efficiency of ejldos and private farms. (b) Case study evidence. The Laguna cotton-producing region of Coahulla state was the site of many pre-1970 case studies. Owing to the relatively generous resource endowment of collective ejidos In this zone, It provided a "best-case scenario' for evaluating the economic efficiency of this sector. Since the Mexican 52 government sought to promote cotton productlon on the collective ejidos, In La Laguna ejldatarlos enjoyed akmost as much access to credit as private farmers (Eckstein et al,1978:61). Many of these studles were conducted under the auspices of the Centro de Investigaciones Agrarlas, a government dependency with an kuplicit brief to defend the land reform (Reyes Osorlo et al,1974; Ecksteln,1 966). 115. A survey conducted In 1967 (Eckstein et al,1978:60-84) sought to distinguish the effects of different forms of land tenure and farm scale on various measures of productivity, using a sample of 208, consisting of small and large private farmers, Individual ejido farmers and collective ejido farmers. The collective ejidos were divided Into three groups according to the Ejldo Bank's evaluation of their creditworthiness: "good" (15 percent of all collective ejidos), "regular" (70 percent) and "bad" (15 percent). Cotton yields per hectare were found to be highest on large private farms, followed In descending order by "good" ejidos, small private farms, Individual ejidos, "regular" ejidos and "bad" ejldos (Table 19). 116. The measurements of gross productivity (value of output per unit of input expenditure) present a complex picture. The "good" ejidos rank the highest In terms of land and total factor productivity; the large private farms rank highest In terms of labour productWity; and the Individual ejidos occupy first place In terms of capital productivity. The least creditworthy collective ejidos (the groups "regular" and "bad") have the poorest overall record. It Is hard to draw any fIrm conclusion from these data because they do not control very carefully for variations In farm scale; this Is a significant ommission because In Mexico there Is some evidence that delivery of subsidized capital Inputs has been blassed In favour of large farms (Cornia,1985). 53 117. However, Eckstein made a further study of ejldo efficiency, this time controlling rigorously for scale (Eckstein et al,1978; Table 20 In this report). He surveyed 313 relatively large, Irrigated farms In nine regions of Mexico, ordered by enterprise size: size Is defined In terms of the gross value of output, rather than the area owned or controlled. To begin with, the productivity of all factors increases In stop with the scale of the enterprise, an Interesting finding because It runs counter to much of the empirical evidence about the efficiency attributes of smali-scale agriculture In the developing natlons (Berry & Cline,1979); Mexico's distinctiveness In this respect has been remarked upon by others (Cornia,1985; Johnston,1 983:230-231). 118. Second, within each of the scale categories, comparing output value with Input value, the ejido performs better than the private farm: the gap Is greatest In the case of capital productivity (confirming Interpretations based on the census data) and total factor productivity; the superiority of the ejido sector Is less marked In the areas of land and labour productivity. These results are very Important because they demonstrate that, In Irrigated regions, enterprise scale overshadows tenure category as the determining factor in economic efficiency. 119. The picture that emerges from these data "seems to be strongly contrary to trends In the rain-fed areas, where ejidos show lower efficiency than private farmers of similar size" (Eckstein et al,1978:83-85). The tentative explanation offered Is that, in the irrigation districts, ejidos had equivalent access to credit and Inputs whereas In rainfed areas they had little access to those resources; Johnston (1983:237) concurs In this view. Throughout the period 1940- 70 governments considered large, irrigated farms to be the most appropriate vehicle for rapid agricultural growth (WIonczek,1982); small producers In the rainfed regions were generally starved of resources and, given the specific constralnts 54 on credit access confronting the ejldos, ejidatarlos probably fared marginally worse than small private farmers. (c) Estimating post-1970 trends. By the early 1970s, agricultural growth had slowed to the point where It was inferlor to the rate of growth of the population; food kuports began to rise sharply (Heath,1988; YfLnez Naudo,1988). Sbnltaneously, thero was an outbreak of widepread rural unrest. Th1 Mexican government responded to these developments by Increasing the flow of subsidized state credit to the raln-fed reglons, Introducing a marketing-subsidies programme for the ejidos (PACE) and placing more emphasis on small-scale irrigatlon schemes (many of these located on the previously-neglectod altiplano of central Mexico). Data on regional shares of national crop output value suggest that, by the mid- 19809, the central and southern regions (where most ejldos are concentrated) were making a much larger contribution than they had done in the early 1960s: betwoen 1949-51 and 1960-62 the Centre's share In crop revenue fell from 30 percent to 25 percent; however, between 1980-62 and 1983, the Centre's share rose from 25 percent to 35 percent (Venezian & Gamble,1967; INEGI,1985). 120. State subsidies to tractor production led to a significant growth In tractor use after 1970; one study Indicates that the adoption rate was higher for the ejido sector than for private farms (SPP-UNIDO,1986). While there are no data In the 1981 Agricultural Cansus on the number of tractors by tenure sector It Is possible to Infer the kIportance of tractor use from the proportion of the cattle herd employed as work anknals (for ploughing and carting). In 1970, 12 percent of cattle In ejidos were used as work animals, compared to 4 percent In the private sector; by 1981, the proportlons had decilned respectively to 5 percent and 2 percent; moreover, on farms up to five hectares, the proportion was the same In 55 1981 (9 percentXTabie 18). Therefore, there Is good reason to suppose that abour productivity on Individual ejidos has now caught up with that of small private farms (altering the pro-1970 picture when, according to the census data, ejidos showed relatively low labour productivity). 121. Evidence Is also avaliable to demonstrate that individual ejidos now have better access to irrigation than they did before 1970: according to the 1981 cens, on farms up to five hectares, 18 percent of ojido cropland--Is irrigated, compared to 13 percent In the case of private farms (Table 11). This suggests that Individual ejidos are no longer likely to have lower crop yields than small private farms. In the case of maize (the most widely-grown crop), the proportion of the sown area under Irrigation was roughly the same for each tenure sector In 1981. Also, In terms of the proportion of the sown area actually harvested (a rough guide to the extent of crop loss), ejido farms fared somewhat better than private farms In 1981 (Table 14). Other land use data from the 1981 census also suggest that Individual ejido farms are now unlikely to be less productive than comparably-sized private farms: the former show a higher proportion of land In agricultural use; and the ratio of fallow land to cropland Is the same for both tenure categories on farms up to five hectares (Table 9). This however may be due to the higher proportion of Irrigated land within the ejido sector. X. CONCLUSIONS AND POLICY RECOMMENDATIONS 122. In contemplating changes to Mexico's land tenure regkne It Is important to consider political as well as economic constraints: specifically, It may be unwise to seek to undo the land reform if the economic payoff from such an Initiative Is likely to be small. Since the mid-1970s, land reform has not been a significant Item 56 on the Mexican government's policy agenda. However, wlthin the government and outside It, the ejido continues to occupy price of place as a symbol of the Mexican Revolution and of the governmont's commitment to the rural poor. Attempts by the past two admlnistratlons to facilitate private sector Investment In the ejldo triggered widespread expressions of dissent. If the prosent government sought to make major alterations to Mexico's land reform institutins It might encounter significant resistance; if such resistance were to place In jeopardy the present admnhlstratlon's programme of economic liberalization, there might well be net losses from the attempt to undo the land reform. 123. More Importantly, the evidence reviewed here suggests that there is no major difference In the productivity of the ejido and private farm sectors. The constraints on agricultural growth In Mexico have been more a function of price policy and the nature of subsidy-allocation, constraints that are felt more or less equally by small farmers In the ejido and private sectors. Rather than deslgntng programmes speciflcally for the ejido sector the major emphasis should remain on Improving Incentives for farmers generally; In particular, It Is Important to ensure that the government's price and subsidy policy does not discriminate against the rural sector. 124. While the institutional differences between Individual ejidos and small private farms appear not to generate major differences In indlcators related to productivity, It Is likely that ejidatarlos face more constralnts on productivity growth than private farmers. Wlthout abollshing the ejido It Is possible to substantially deregulate the sector: this entails making plecemeal Improvements to the existing institutlonal structure of land reform. Specifically, 57 1. The law about the size of ejido parcels has proved unenforceable and should be abandoned; granting of titles to ejidatarlos should not be contingent on the size of their parcel. 2. The existing drive to provide ejldatarlos wlth titles to their parcels should be accelerated, In order to provide them with maxinum socurity of tenure; 3. For private farms, holding sizo colliigs should only vary according to land quality, not according to the crop grown; restrictions on the maxinknum size of private cattle ranches should be clarified; the proportion of these ranches that may be planted to crops should be made clear; If the land reform law Is altered In this respect there will be less basis for arbitrary expropriation, favouring higher output growth and on-farm Investment In the private sector. 4. EJldatarlos should be free to enter Into leasing or sharecrop arrangements with other ejldatarlos or with private farmers, without facing the risk of having their parcels confiscated; renting and sharecropping may also create employment wlthin the ejido (since land-deficit families that do not migrate are able to work the land of those who have migrated); although, In many cases, renting and sharecropping Increase wealth concentration, they may In some cases have positive distributional effects; In any event, they ensure efficient exploitation of ejido land by placing It in the hands of those with the means and the vocation to work It. 5. Ejldatarlos should be allowed to sell their parcels back to the ejido: as long as sales do not Involve outsiders the integrity of the ejido will be preserved; the sale price of the parcel should compensate 58 the ejidatarlo for kIprovemonts he has made to It; recommendatlons (4) and (5) are conducive to the formation of larger and more offclont units of enterprise within the ejido sector, counteracting the nogative effects of parcel fragmentation. 6. Restrictions on the use and management of communal ejido lands should be anmnended to allow leasing, private Investment and joint ventures; government will need to ftid an appropriate way of onforchg standards concernhg timber cutting, replanting rates and pastoral carrying capacity, In order to avoid environmental degradation: It Is essential that the qxternal costs entailed by expWltation of communal lands be Internalized" (i.e. borne by lndividual ejidatarlos and private Investors), In order to promote ecologically-responsible forestry and stock-rearing activities. 7. Ejldatarlos should be permitted to make Individual or voluntary group credit applications to Banrural; the evaluation of Ioan applications should be based on creditworthiness criteria; S. The burden of default should be borne only by the individuals or groups responsible for defaults, not by the whole ejido; 9. All Banrural credit should be provWed In cash with full responsibility for Input purchase and delivery being borne by the farmer; farmers should be free to plant whatever crops they wish with the credit they recelve. GRAPH 1. MAIIZE PRICEI' AND WVAGE ( TENDS', 1.940-8. cr 0%~ ~ 4 t 4 0 - to4 0-- _ _ _ _ _ _ _ _ /'A 130 /AMAIZ 120 q 70 - 60- 50 - 100- 40 0 > / T--,,,-T-r-T---T--TT-T-1 T wWT IT - 40 45 50 55 60 6.S 70 75 flO R6 ) F.lRr. 9rs- sot986); 60 TABLE 1: VARIATION IN THE INTENSITY OF LAND REFORM BY PRESIDENTIAL TERM, 1900-1988. Period Land redistributed Percent of total (Hectares) area redistributed by 1988 1900-1914 194,495 0.2 1915-1934 11,543,833 11.0 1935-1940 18,786.131 17.9 1941-1946 7,277.697 6.9 1947-1952 4,590,381 4.4 1953-1958 6,056,773 5.8 1959-1964 8,870,430 8.4 1965-1970 24,738,199 23.5 1971-1976 12,773,888 12.1 1977-1982 6,097,005 5.8 1983-1988 4,167,588 . 4.0 TOTAL 105,167,588 4.0 SOURCE: 1900-1982: SPP-INEGI, Estadisticos Hist6ricos de Mdxico, Tomo I, Mexico D.F., SPP, 1986.1983-1988: Sexto Informe Presidenclal, 1988 (Anexo Estadistico), Mexico D.F., SPP, 1988. 61 TABLE 2: Proportion of Farms In the Ejido Sector by State and Region, 1981* States and Regions EJldo/Total % All of Mexico 63.8 North Pacific 77.0 Sonora 47.0 Slnaloa 81.0 California Norte 69.0 Baja Calfornia Sur 23.9 Nayarft 92.8 North 68.9 Chihuahua 65.3 Coahulla 78.1 Nuevo Leon 35.7 Tamaulipas 65.0 San Luis Potosi 80.6 Durango 83.0 Zacatecas 62.8 Center 56.8 Jalisco 56.2 Aguascallentes 72.2 Guanajuato 65.0 Queretaro 69.5 Hidalgo 61.2 Talaxcala 8.6 Morelos 80.0 Puebla 48.0 Michoacan 72.0 Mexico City 46.3 Mexico 48.6 South Pacific 66.9 Coikna 82.4 Guerrero 74.6 oaxaca 58.4 Chlapas 73.1 Gulf 64.5 Veracruz 66.8 Tabasco 42.7 Campeche 81.9 Yucatan 72.3 Quintana Roo 87.9 Source: Agricultural Census, 1981. *Refers to number of enterprises, not area occupied. 62 TABLE 3: Extent of Land Reform by Region, 1985 Land Distributed Total Area (1)/(2) (hectares) (hectares) (X) All MexIco 102,876,920 196,718,300 52.3 North Pacific 18,430,252 41,443,700 44.5 North 38,131,949 80,057,800 47.6 South Pacific 12,344,546 23,850,000 51.8 Centre 13,041,635 27,466,900 47.5 Gulf 12,359,717 23,899.900 51.7 Note: See Table 2 for list of states comprising each region. Source: INEGI, Estadisticos Hisoricos de Mexico, Tomo I, Mexico D.F., SPP, 1986; Statistical Abstract of Latin America. Vol. 2C. Table 301. ______ ~~~~~~~~63 TABLE 4: Dlstributlon of Farm Enterprises by Tenure Category, 1981. (a) Number of farms Private Ejido Mixed Total All farmn 1,003,374 2,099,038 189,668 3,292,100 Farms of 5 hectares or less 565,846 1,220,058 120,824 1,906,728 (b) Percent of farms by tenure category Private Ejldo Mixed Total All farms 30.5 63.8 5.7 100.0 Farms of 5 hectares or less 29.7 64.0 6.3 100.0 (c) Percent of farms In each tenure category that occupy five hectares or less Private Ejido MIxed Total 56.4 58.1 63.7 57.9 SOURCE: VI Censos Agricola-Ganadero y EJidal,1981, (Cuadro 01) INEGI-SPP: Mexico D.F., 1988. 64 TABLE 5: Proportion of Land not In Production by Tenure Category, 1981. Private Ejldo Mixed '000 has % '000 has X '000 has X ALL ENTERPRISES In Production 54,199 73.4 12,975 85.2 2,115 73.1 Not In Pro- duction 19,662 26.6 2,260 14.8 778 26.9 Total 73,861 100.0 15,235 100.0 2,893 100.0 ENTERPRISES UP TO FIVE HECTARES In Production 919 93.6 2,778 94.1 259 98.1 Not in Pro- duction 63 6.4 174 5.9 5 1.9 Total 982 100.0 2,952 100.0 264 100.0 Note "Not In production" means that the land was not put to agricultural or silvicultural use in the perlod April-September 1981; this category Includes enterprises that have never used the land for these purposes. SOURCE: Same as Table 4 65 TABLE 6: Estkmates of Area of the Ejido Sector DATE/ SOURCE 1970 1970 1981 1982 198S Census SRA Census SRA SRA | '~~000 Hectares (a) CROPLAND 12,763 12,511 10,012 14,230 14,380 - rainfed 10,993 10,831 8,339 12,360 12,471 - irrigated 1,760 1,680 1,673 1,870 1,909 (b) TOTAL AREA I (1) 89,724 82,148 na 101,319 102,877 (2) na na 16,235 na na Key to (b) (1) - Whole ejido sector; (2) ' Sum of parcels in individual ejidos; ndividual ejldos; na - Not available. SRA - Secretarla de Reforma Agraria Note: The census data for "Total area" are not comparable owing to the different way in which ejido land was measured In 1970 and 1981; the degree of comparability is greater In the case of 'cropland' (see Appendix A). wTotal area" covers cropland, forest, natural grazing and unproductive land. "Cropiand" includes sown pasture. SOURCE: Agricultural Census, 1970 & 1981; INEGI, 1986. 66 TABLE 7: Number of Farm Enterprises by Tenure Category 1981 Prtvate Ejldo Mixed '000 % '000 % '000 % farms farms farms Holding size (has) 0-2 398 39.7 647 30.8 66 34.9 3-5 168 16.7 573 27.3 55 29.1 6-20 213 21.2 809 38.5 55 29.1 21-50 101 10.1 58 2.8 8 4.2 51-100 54 5.4 4 0.2 2 1.1 100-1000 58 5.8 9 0.4 3 1.6 1001-2500 6 0.6 * * ' * Over 2500 5 0.5 - - Total 1003 100.0 2100 100.0 189 100.0 SOURCE: Same as Table 1. 67 TABLE 8: States Where More 25% of Farms were "not In Productlon" In 1981. Share of land In holdings corresponding to farms not In production (%) Private EJldo State Baja California 94.7 47.1 Baja California Sur 82.9 7.7 Nayarit na 33.9 Slnaloa 31.2 na Sonora 42.7 54.9 Coahuila 38.3 28.9 Nuevo Leon na 26.3 Colima 56.2 na Guerrero 32.9 36.9 Oaxaca 30.7 na Campeche 78.8 na Quintana Roo 97.3 72.8 Yucatan 29.0 na na-Share of farms not In production less than 25%. SOURCE: Same as Table 4. 68 TABLE 9: Land use by Tenure Category, 1981. Private EJldo MIxed '000 hectares TOTAL 73,862 15,236 2,892 CuWtivated 8,735 8,279 1,388 Fallow 1,875 1,733 197 Natural Pasture 27,427 2,154 580 Forest 28,428 2,588 611 Other 9,699 481 138 TOTAL 100.00 100.0 100.00 Cultivated 11.8 54.3 47.9 Fallow 2.3 11.4 8.8 Natural Pasture 37.1 14.1 19.4 Forest 35.8 17.0 21.1 Other 13.0 3.2 4.8 ENTERPRISES UP TO FIVE HECTARES '000 hectares TOTAL 982 2,952 265 Cultivated 699 2,294 221 Fallow 69 259 16 Natural Pasture 93 189 12 Forest 80 172 11 Other 41 38 5 TOTAL 100.0 100.0 100.0 Cultivated 71.2 77.7 83.4 Fallow 7.0 8.8 8.0 Natural Pasture 9.5 6.4 4.5 Forest 8.1 5.8 4.2 Other 4.2 1.3 1.9 Note: "EJido" refers to the aggregate of Individual parcels, not to collective ejidos or communally-worked areas of parcelilzed ejldos; "Cultivated" Includes planted pasture; "Other" refers prknarily to land that Is not being farmed and/or Is Inappropriate for agriculture. SOURCE: Same as Table 4. 69 TABLE 10: Number of Cattle by Tenure Category 1970 & 1981 Private EJldo Mixed '000 Head ALL. FARMS Total Herd 1970 13,003 4,734 na 1981 14,067 7,613 823 Purebred Cattle 1970 1,995 160 na 1981 3,163 620 97 FARMS UP TO 5 HECTARES Total Herd 1970 2,633 na na 1981 855 2,378 187 Purebred Cattle 1970 124 na na 1981 88 189 16 SOURCE: Agricultural Census, 1970 & 1981. 70 TABLE 11: Proportion of Cropland that Is Irrigated by Tenure Category Private Ejldo Mixed '000 has % '000 has ' 000 has X ALL ENTERPRISES Irrigated 1,823 16.6 1,878 17.7 219 13B Unirrlgated 9,133 83.4 8,729 82.3 1,393 86A Total 10,956 100.0 10,607 100.0 1,612 100D ENTERPRISES UP TO FIVE HECTARES Irrigated 104 13.1 464 17.7 38 158 Unirrigated 687 86.9 2,152 82.3 202 84 Total 791 100.0 2616 100.0 240 1OOD SOURCE: Same as Table 4. 71 TABLE 12: Proportlon of Cropland under Irrigation by Region and Tenure Category, 1970 & 1981 Percentages Ejldo Private 1970 1981 1970 1981 All Mexico 13.8 17.7 15.5 186. North Pacific 59.4 45.5 68.8 59.7 North 20.1 17.9 26.7 18.3 South Pacific 3.6 4.0 4.1 4.5 Centre 13.0 21.2 15.6 13.3 Gulf 1.7 2.5 1.1 2.3 Note: For list of states comprising each region see Table 2. SOURCE: Agricultural Census, 1970 & 1981 72 TABLE 13: Main Crops Cultivated In Ejldo and Private Sector, 1981. Private Ejldo Area Harvested* 5.010 (100.0) 7,203 (100.0) ('000 has.) Maize 2,456 (49.0) 4,274 (59.3) Beans 738 (14.7) 1,175 (16.3) Sorghum 583 (11.6) 581 (8.1) Wheat 341 (6.8) 271 (3.8) Safflower 202 (4.0) 168 (2.3) Soybean 172 (3.4) 166 (2.3) Other** 518 (10.5) 568 (7.9) 3 Main crops only (Sum of autumn-winter 1980-81 and spring-summer 1981 crop cycles) * Sesame seed, cotton, rice, oats, barley, chickpea, chile pepper, tomato, tobacco. SOURCE: Same as Table 4. 73 TABLE 14: Maize: Relation of Sown to Harvested Area Private Ejldo Mixed Maize: 1980-1981* (1) Sown area ('000 has) 2,933 4,813 721 (2) Harvested area ('000 has) 2,456 4,273 - 645 (3) Percent of (2) under irrigation 13.9 14.9 21.7 (4) (2)/(1) (X) 83.7 88.8 89.5 * Refers to the sum of two crop cycles: Autumn-Winter (1980-81) and Spring-Summer (1981) SOURCE: Same as Table 4. 74 TABLE 15: Number of Farm Enterprises engaged In Cooperative forms of Production Private Ejido Mixed '000 '000 '000 farms X farms % farm % All enterprises 921 100.0 1958 100.0 184 100.0 Enterprises working cooperatively - within kin group 9 1.0 11 0.6 2 1.1 - between kin group 2 0.2 114 5.8 9 4.9 Note "Enterprises working cooperatively" does not Include collective ejidos: It refers to all forms of work association between different households, both within the same kin group and between different families, where work Is not remunerated by payment of a wage. SOURCE: Same as Table 4. 75 TABLE 16: Proportion of Cattle Herd used as Work Anknals (1) (2) (3) All cattle Work anknals (2)/(1) ('000 head) ('000 head) (X) ALL FARMS Private 1970 13,003 549 4.2 1981 14,067 287 2.0 Ejldo 1970 4,735 562 11.9 1981 7,613 406 5.3 FARMS UP TO 5 HECTARES Private 1970 2,633 258 9.8 1981 855 82 9.6 Ejido 1970 na na na 1981 2,378 222 9.3 SOURCE: Agricultural Census, 1970 & 1981. 76 TABLE 17: Trends In the Volume of Agricultural Credit, 1983-88 Millions of 1980 Pesos (1) (2) (3) (4) BANRURAL BANRURAL FIRA-SNC TOTAL EJldatarlos 1983 42,298 30,020 67,342 109,640 1984 49,489 36.262 77.411 126.900 1985 53,831 40,018 94,616 148,447 1986 50,423 37,913 69,783 120,206 1987 42,045 30,274 49,213 91,258 1988 62,760 46,483 75,352 138,112 Share of Banrural Share of Share of Banrural credit in total ejidatarlos In credit to ejidatarlos in Agricultural credit all Banrural credit total agricultural credit (1)J(4) (%) (2)/(1)(%) (2)/(4)(%) 1983 38.6 71.0 27.4 1984 33.4 73.3 24.5 1985 36.3 74.3 27.0 1986 41.9 75.2 31.5 1987 46.1 72.0 33.2 1988 45.4 74.1 33.7 SOURCE: Banrural, Informe de Autoevaktacion Sexenal, 1982-1988, Mexico D.F., 1989. 77 TABLE 18: Agricultural Credit Trends - 1980-88 8ANRURAL A FIRA: Annual Lending and Portfolio (In billions of 1988 Mex8) BANRURAL 1980 1981 1982 1983 1984 1986 1988 1987 1988 Balances at year-end 1/ 6,829 6,641 4,441 3,663 3,922 4,380 3,329 2,774 4,036 I. Working Capital Loans 2,806 3,416 2,209 1,860 2,282 2,847 2,149 1,988 2,996 II. Investment Loans 2,824 3,228 2,233 1,702 1,641 1,533 1,180 788 1,039 LOANS GRANTED DURING THE YEAR V 6,337 6,996 4,802 3,806 4,314 4,093 4.076 8,710 3,943 I. Working Capital Loans 4,060 4,843 3,004 3,061 3,636 4,023 3,499 3,270 3,422 II. Investment Loans 1,287 1,362 998 744 779 070 676 434 622 FIRA Balances at year-end / 3,667 3,616 2,710 2,687 2,983 3,262 2,134 1,886 2,889 I. Working Capital Loans 1,766 1,976 1,811 1,729 2,066 2,113 1,624 1,262 1,732 II. Investment Loans 1,881 1,641 1,100 867 928 1,139 010 632 968 NEW DISCOUNTS DURING THE YEAR V 3,610 3,679 3,464 2,962 3,274 3,700 2,686 2,882 2,689 I. Working Capital Loans 1,727 2,010 2,062 1,973 2,266 2,407 1,918 1,762 1,732 II. Investment Loans 1,883 1,6d9 1,401 978 2,019 1,208 709 890 968 BANRURAL & FIRA Balances at Year-end 9,196 10,167 7,161 6,139 6,905 7,632 6,403 4,869 8,724 I. Working Capital Loans 4,612 6,390 3,819 3,680 4,336 4,960 3,873 3,240 4,728 II. Investment Loans 4,886 4,767 3,332 2,680 2,689 2.872 1,790 1,419 1,996 NEW LOANS/DISCOUNTS DURING THE YEAR 8,947 9,674 8,260 8,767 7,686 8,398 6,781 6,362 6,632 I. Working Capital Loans 6,777 6,653 6,867 6,036 8,790 6,430 6,418 6,038 6,164 II. Investment Loans 3,170 2,920 2,399 1,722 1,708 1,969 1,343 1,324 1,478 1/ Portfolio balances were inflated according to the end-of-year CPI. 78 TABLE 19: Comparing the Efficlency of Ejidos and Private Farms: (i) A Survey of La Laguna, 1967. individual 1967 Pesos Private farms Ejldatarlos Collective ejidos Large Small Good Average Bad Gross product per hectare 7,804 7,977 4,620 8.540 5.100 3,450 Gross product per unit of capital 0.81 1.31 2.51 1.03 0.80 0.95 Gross product per manday 129 47 43 58 48 31 Total factor productivity 1.15 1.29 1.24 1.35 0.90 0.65 NOTES: (I) N-208 farms; (2) 'Large farms' refers to those over ten hectares; 'small farms' have under ten hectares; (3) 'Good', 'average', 'bad' refers to the Ejido Bank's assessment of the creditworthiness of the collective ejidos In the sample; (4) Figures for private farms and individual ejldatarlos refer to Individual farm operations; figures for the collective ejidos refer to values for the whole ejldo divided by the number of ejldatarlos; (5) 'Gross product' refers to the value of all farm production In 1967: crops, livestock, forestry; Includes changes in inventory; (6) 'Capital' refers to the value of livestock, equipment and plantations, excluding land, but Including half the value of purchased Inputs (land and equipment rent, hired labour, irrigation water, fertilizer, pesticide, Interest on loans) as an estimate of working capital; (7) 'Mandays' excludes hired labour; (8) 'Total factor productivity' Is gross product divided by purchased Inputs plus capital plus an knputed cost for owners' family labour. SOURCE: Ecksteln et al (1978), p. 64. 79 TABLE 20: Comparison of average productivity and related values for Private and Ejido farms of three sizes in a sample from Irrigated lands in Mexico, 1987 Small Farms t/ Medium Farms o/ Large Farms p/ Private Ejido Private Ejido Private Ejido No. of Farm Units 41 122 34 43 78 6 Land Productivity Cultivated land (hectaros) 7.3 6.0 19.0 13.6 129.8 47.6 Product per h-ctsre (pesos) 2400 2500 3200 4800 7734 8069 Gross income b/ hectare (pesos) 673 1267 610 1462 28U6 3813 Capital Productivity Capital per farm (1000 pesos) 22 12 111 27 1040 160 Capital per hectare (1000 pesos) 3.4 2.1 656 2.4 8.3 3.7 Product/capital .68 .93 .47 1.24 .72 1.16 Labor Productivity Man-days per farm S/ e88 468 1284 989 8169 3021 Man-days per hectare c/ 99 82 70 87 66 69 Product per man-day (pesos) E/ 24 30 46 65 119 116 Total Factor Productivity Product per farm (1000 pesos) 16 14 61 64 969 362 Cross income b/ per farm (1000 pesos) 4.2 7.6 11.6 19.7 369 172 Product per total inputs J .89 .98 .80 1.10 1.17 1.44 Product per total Inputs JI .96 1.40 .82 1.27 1.18 1.61 Value-added per total inputs I/ .82 1.05 .46 .80 .89 1.06 */ Sizo groups based on gross value of output in pesos: small = 6,000 - 25,000; medium = 26,000 - 100,000; large 100,000 and over. b/ Value of product minus purchased Inputs. c/ Hired labor plus labor of owner/ojidatarios. a, Inputs include those purchased plus inputed values for land and owner/ejidatario labor. */ Same as d., but excluding imputed value of owner/ojidatario labor from the denominator. 80 APPENDIX A A NOTE ON THE DIFFICULTY OF INTERPRETING THE RESULTS OF THE 1981 AGRICULTURAL CENSUS. The 1981 AgrIcultural Census Is the most up-to-date source of nationwide data which explicitly disaggregates by tenure category. However, compared to the 1970 Census (which has been extensively analysed In the secondary literature) the 1981 Census provides a less than adequatereferent for comparing private and ejldo sectors. There are also a number of reasons why It Is difficult to meaninfully compare the 1970 and 1981 censuses. First, the 1981 data Is highly schematic: the computerized data base containing the census returns was destroyed In the 1985 Mexico Clty earthquake and It was only possible to derive estinates based on a random sample of 10 percent of the original questlonairres. The Information contained In the one document published Is much less complete than that of 1970: for example, there Is no data on crop output making It Impossible to compare yields between the private and ejido sector. Second, the definitlon of holding sizes and land tenure categories Is substantially different for 1970 and 1981. In 1970, private holdings are broken down by size Into only two groups: propertles of five hectares or less and properties larger than five hectares; In 1981, farms (ejldo parcels as well as private holdings) are classified Into eight holding-size strata. More kIportantly, whereas In 1970 data on ejidos ano comunidades agrarlas took as their unit of reference the whole ejido, In 1981 the unit of reference is the Individual parcel within the ejldo or comunidad. This conceptual adjustment reflects the prevailing reallty of the Mexican land reform sector where household enterprises essentially work the land on an Individual basis, although nominally they form part of a community of producers. The 1981 adJustment facilitates comparlson of household enterprises In the ejido and the private sector. However, a major falling of this approach concerns the excluslon of collective ejidos and the communal areas of parcellized ejidos: it Is therefore kmpossible to ascertain the total land base and resource endowment of the ejido sector and Its proportional significance In relation to the private sector. This adjustment Is reflected In the area covered by the census: a mere 91.9 millon hectares In 1981 (roughly 47% of national territory), compared to 169 miilon hectares In 1960 and 139 mililon hectares In 1970. According to the 1981 Census only 17 percent of the land In holdings corresponds to the ejido sector. Statistics compiled by the Agrarian Reform Ministry Indicate that by the end of 1982 51 percent of the natlonal territory had been turned over to the reform sector; this tallies with the 1970 Census results which Indlcated that Just under half of the area surveyed was in ejidos or comunidades. In other words, roughly two-thirds of the ejido resource base Is unaccounted for In the 1981 census. 81 A further difference between the 1970 and 1981 Censuses concerns the Introduction, In 1981, of a new tenure category (Mixta) corresponding to those farms that are operated Jointly by ejidatarlos and private farmers. The Inclusion of this category was undoubtedly Influenced by the Ley de Fomento Agropecuarlo (1980) which gave legal sanction to the association of farm enterprises from the private and land reform sectors. Typically, ejidatarlos contribute land (and possibly labour) while the private farmers provide working capital. It Is not clear how, In legal terms, this arrangement differs from the straightforward renting-out of ejido land, a practice that Is formally proscribed by the Agrarian Reform Law. The renting-out of ejido land has been widespread In Mexico since the earilest days of the land reform. Because It Is Illegal the extent of this practice Is akmost certainly not fully reflected In the new tenure category, mixta. According to the 1981 Census only 6 percent of farm enterprises and 3 percent of the land surveyed Is ciasslfled as mixta. It Is safe to assume that the effective control and management of land In Mexico Is concentrated In fewer hands than the Census data would suggest since large numbers of ejidatarlos clandestinely hand over their land to other producers In exchange for a money rent, thus exercising no Influence over the use to which that land Is put. 82 BIBLIOGRAPHY Adams D.W. & R.C. Vogel (1988) "Rural financial markets In low-income countries", World Development, April. Arlzpe, L. (1981) "Relay migratIon and the survival of the peasant household" In J. Balan (ed) Why People Move: Internal Mlgratlon and Development, Paris: UNESCO. Bailey, J.J.(1981) "Agrarian reform In Mexico" Current History, Vol. 80, No. 469, November. Banrural (1989) Credit data suppiled to Graclela Lltuma, Country Department II, World Bank. Barkin, D. (1977) "Desarrolio reglonal y reorganizacl6n campesina: La Chontalpa como reflejo del gran problema agropecuarlo mexlcano" Comerclo Exterior, Vol. 27, No 12. Bartra R. (1975) "Peasants and political power In Mexico" Latin American Perspectives, Vol. 2, No. 2. Berry, R.A. & W.R. Cline (1979) Agrarian Structure and Productivity In Developing Countries, Baltimore: Johns Hopkins University Press. Boruconsa (1982) Memoria: Programa de Apoyo a la Comercializaci6n Rural, 1976-1982 Mexico City: Bodegas Rurales Conasupo, Gerencia de Programas Especiales. Brannon, J. & E.N. Baklanoff (1984) "The polltical economy of agrarlan reform In YucatAn, Mexico" World Development Vol. 12, Nos. 11-12. Brannon, J. & E.N. Baklanoff (1987) Agrarian Reform & Public Enterprise In Mexico: The Political Economy of Yucatan's Henequen Industry. Tuscaloosa, Alabama: University of Alabama Press. CIMMYT (1977) The Puebla Project: Seven Years of Experlence (1969-73), Mexico City: CIMMYT. C6rdova, A. (1974) La polltica de masas del Cardenismo, Mexico City: Era. Cornia, G.A. (1985) "Farm size, land yields and the agricultural production function: An analysis for 15 developing countries" World Development, Vol. 13, No 4. Craig, A. (1983) The First Agraristas, Berkeley: University of California. De Wait, B.R. (1979) Modernization In a Mexican Ejldo: A Study In Economic Adaptation, Cambrldge: Cambridge University Press. Diaz-Cisneros, H. (1983) The Agrarian Reform and Rural Development In Mexico, (Development Paper No. 149) Cambridge, Massachusetts: Harvard Institute for International Development. Dovring, F. (1970) "Land reform and productivity In Mexico," Land Economics, No. 3. 83 Ecksteln, S. (1966) El ejldo colectivo en Mexico. Mexico City: Fondo de Cultura Econ6mica. Eckstein, S. et al (1978) Land Reform In Latin America: Bolivia, Chile, Peru and Venezuela, World Bank Staff Working Papers, No. 275, Washington D.C. Exc6lslor (Mexico City), Manuel Buendla column, Sectlon A, p. 1, 19th & 20th March, 1984. Finkler, K. (1978) "From sharecroppers to entrepreneurs: peasant household production strategies under the ejido System of Mexico" Economic Development and Cultural Change Vol 27, No. 1. Flores, E. (1959) "The structure of land use changes In the economic development of Mexico," Land Economics. Vol. 25, May. Flores, E. (1969) "From land reform to Industrlal revolution: the Mexican case" Developing Economles Vol. 7, No. 1, 1969. Fox, J.A. (1986) The Polltical Dynamics of Reform: The Case of the Mexican Food System, Unpublished PhD Dissertatlon, Massachusetts Institute of Technology. Friedrich, Paul (1970) Agrarian Revolt In a Mexican Village. Englewood Cliffs, NJ: Prentice-Hall. G6mez Oliver, L. (1984) El Desarrollo Agropecuarlo de M4xico: Pasado y Perspectivas: (Tomo 7) El Ingreso y su Distrlbuci6n Mexico City: SARH/CEPAL. Goodman, L.W. et al, (1985) Mexican Agriculture: Rural Crisis and Policy Response, Working Paper No. 168, Latin American Program, Wilson Center, Washington D.C. Gregory, P. (1986) The Myth of Market Failure, Baltimore: Johns Hopkins University Press. Hansen, R.D. (1971) The Politics of Mexican Development, Baltimore: Johns Hopkins University Press. Heath, J.R. (1985) "Contradictions in current Mexican food pollcy" in G. Phiilp (ed) Politics in Mexico, London: Croom Helm. Heath, J.R. (1987) "Constraints on peasant maize production: a case study from Michoacan" Mexican Studles, Summer. Heath, J.R. (1988) "An overview of the Mexican agricultural crlsis" In G. Philip (ed) The Mexican Economy, London: Routledge. Hewitt de AlcAntara, C. (1978) La modernizaci6n de la agricultura mexicana, 1940-1970 Mexico Clty: Sg1go XXI. Hewitt de Alchntara, C. (1980) "Land reform, peasant livelihood and power In rural Mexico" in D.A. Preston (ed) In D.A. Preston (ed) Environment, Society and Rural Change In Latin America, Chichester: John Wiley. INEGI (1985) Anuarlo Estadistico de los Estados Unidos Mexicanos Mexico Clty: SPP. 84 INEGI (1986) Estadisticos Hist6ricos de M4xico Vol 1, Mexico Clty: SPP. INEGI-SPP (1988) VI Censos Agricola-Ganadero y EJldal, 1981 Mexico City: SPP. Johnston, B. (1983) "The design and redeslgn of strategies for agricultural development: Mexico's experience revisited" In C.W. Reynolds & C. Telio (eds) U.S.-Mexico Relatlons: Economic and Social Aspects, Stanford: Stanford University Press. Johnston, B. & P. Kilby (1975) Agriculture and Structural Transformation New York: Oxford University Press. Montaftez, C. (1989) Interview with author, Secretaria T4cnica, Banrural, Mexico City (August). Mueller, M.W. (1970) "Changhig patterns of agricultural output and productMvity hI the private and land reform sectors In Mexico, 1940-60" Economic Development and Cultural Change, Vol. 18, No. 2. Mummert, G. (1987) "The transformation of the forms of social organization In a Mexican ejido (1924-81)" International Social Science Journal, Vol. 39, No. 4. NACLA (1976) "Harvest of anger" Latin America & Empire Report Vol. 10, July-August. Nguyen, D.T. and M.L. Martinez Saldivar (1979) "The effects of land reform on agricultural reform, employment and Income distribution: a statistical study of Mexican states" The Economic Journal. Vol. 89 September. Norton, R.D. (1987) "Policy choices In agricultural trade between Mexico and the United States" In B. Johnston et al (eds) U.S.-Mexico Relations: Agriculture and Rural Development Stanford: Stanford University Press. Pillet, B. (1973) "Land tenure and agricultural productivity In Mexico" Unpublished ms, Food Research Institute, Stanford University, March. Redclift, M.R. (1983) "Production programs for small farmers: Plan Puebla as myth and reality" Economic Development and Cultural Change. Vol. 30, No. 3. Rello, F. (1987) State and Peasantry In Mexico: A Case Study of Rural Credit In La Laguna, Geneva: UNRISO. Reyes Osorlo, S. et al (1974) Estructura agraria y desarrollo agricola en Mexico. Mexico City: Fondo de Cultura Econ6mica. Roberts, K.D. (1982) "Agrarian structure and labor mobility In rural Mexico" Population and Development Review. Vol. 8, June. Sanderson, S. (1984) Land Reform In Mexico: 1910-1980. Orlando: Academic Press. Soberon-Ferrer, H. (1986) Land Reform and Economic Efficiency: The Case of Puebla, Mexico Unpublished PhD dissertation, Clemson University. 85 SPP-UNIDO, Blenes de capital e Insumos para la agricultura mexicana, Mexico Clty: Naclonal Financiera. Stavenhagen, R. et al (1968) Neolatifundlsmo y explotacl6n Mexico City: Nuestro Tiempo. Stavenhagen, R. (1975) "Collective agriculture and capitalism In Mexico: a way out or a dead end?" Latin American Perspectives, No. 5, Sunmmer. Stringer, H. (1972) "Land, farmers and sugar cane In Morelos, Mexico" Land Economics Vol 48, No 3. Suhrez, B. (1982-83) "Las semillas, el estado y las transnaclonales" Problemas del Desarrollo Vol 13, Nos. 51-52. Van Ginneken, W. (1980) Socioeconomic Groups and Income Distributlon In Mexico New York: St Martins. Venezian, E. & L. Gamble (1967) "El desarrollo de la agricultura mexicana" Investigaci6n Econ6mica January-June. Villa Issa, M. (1977) "El mercado de trabaJo y la adopcion de tecnologia nueva de produccion agricola", Chapingo: Coleglo de Posgraduados. Warman, A. (1977) "El neolatifundismo mexicano: expansl6n y crisis de una forma de dominio" Comerclo Exterior Vol. 25, No. 12. Weckstein, R.S. (1970) "Evaluating Mexican land reform" Economic Development and Cultural Change Vol. 18, No. 3. Wilkie, R. (1971) San Miguel: A Mexican Collective Ejldo. Stanford: Stanford University Press. Wionczek, M. (1982) "The roots of the Mexican agricultural crisis: water resource development policies (1920-1970)," Development and Change Vol. 13, No. 3. Womack Jr., J. (1969) Zapata and the Mexican Revolution, New York: Knopf. World Bank (1989) Mexico: Agricultural Sector Report, Latin America & Caribbean Region, World Bank, Washington D.C., January 24. Yates, P.L. (1981a) Mexico's Agricultural Dilemma. Tucson: University of Arizona Press. Yates, P.L. (1981b) "Mexican land reform, 1959-69: A comment" The Economic Journal Vol 91, September. Yinez Naude, A. (1988) La agricultura mexicana en crisis: Reflexiones te6ricas y anAllsis empirico Mexico Clty: Fondo de Cultura Econ6mica. Zaragoza, J.L. & R.Macias (1980) El desarrollo agrario de Mexico y su marco juridico. Mexico City: Centro Naclonal de Investigaciones Agrarlas.