see the new book from Stanford University Press

Acknowledgements

The question of the political nature of informal economic activity is theoretically important from a variety of viewpoints. From the standpoint social movement theory, organization among informal economic actors shows grassroots, non-traditional forms of organization among groups that were hitherto believed to be politically as well as politically marginal, helping to identify strategies that such relatively powerless groups use to thwart state attempts to control or eliminate their activity (Parnell 1992). In the area of economic theory, it highlights the political nature of the modern economic system--formal or informal. In the area of political theory, it can potentially focus attention on the nature of the state itself, since informal economic activities are often allowed to continue at least partly based on not only corruption and inefficiency but also on basic policy cleavages within the state itself as well as administrative choices about how best to carry such policy out (Fernandez-Kelly & Garcia 1989).

From a practical point of view, the nature of organization and resistance within the informal economy also suggests avenues not only for the emergence of effective grassroots interest groups among some of the poorest members of third-world societies, but also the potential that policy choices which promote development from below can be put forward by the very sectors that development planners have most desired to reach. Assuming that such groups can articulate their interests better than policy planners who approach development from above, this may lead to more realistic processes of development in such societies, and may even open up new opportunities for growth even in first world nations where substantial pockets of poverty continue to exist in the face of relative luxury.

In this paper, I approach these issues by a historical comparison of two periods of street vendor repression in Mexico City. It is argued that high levels of organization among this informal economic groups, even though based upon the attempt of the state to control street vendors for political purposes, have been largely successful at defending at least one realm of interest central to the nature of street vending itself: access to public space.

"Hegel remarks somewhere that all facts and personages of great importance in world history occur, as it were, twice. He forgot to add: the first time as tragedy, the second as farce."

--Karl Marx, The Eighteenth Brumaire of Louis Bonapart (1963)

The government of Mexico City, the Department of the Federal District (DDF), is in the midst of one of the most politically sensitive operations of recent years: the relocation of approximately 10,000 ambulatory vendors (street vendors) from the streets of the Historical Center into almost 40 market buildings. The process has been long and difficult. Planning in a broad sense dates back more than a decade, reflected in numerous studies over that period on the type, quantity and location of vendors. The 1985 earthquake sparked a search by several agencies to see whether vendors would fit into urban lots made available by the collapse or condemnation of earthquake-damaged buildings. But real planning did not begin until after the 1991 elections, when a strong showing for the PRI gave the government a relatively free hand with vendors who had been among their most loyal political supporters during the economic and political crisis of the 1980s.(1)

This recent process seems all the more remarkable in the face of the consistent failure of the city's attempts to control street vending in the last decades as the number of street vending has grown despite dozens of administrative proposals to limit or reverse that growth. Even the refusal of the city to grant new permits to street vendors since 1984 and the formation of the Coordinaci¢n de Abasto Popular (COABASTO) the following year to organize and reduce the phenomena did not prevent the number of vendors from doubling over the subsequent decade, usually with the knowledge and even the support of elected and administrative officials.(2)

The difficulty of controlling vending has led many observers to claim that street vendors are a "mafia" that is able to manipulate officials to do their bidding (Baca 1990), and many officials admit in private that street vendor leaders are so well organized politically that it is extremely difficult for city agencies operating alone to take any effective action against them. Street vendor leaders proudly boast of their relations with high level politicians and officials, while some have themselves been elected to office or are invited with other local dignitaries to presidential receptions.(3) The power of leaders is based upon the fact that, possibly more than any other city in the world, Mexico City's vendors are extraordinarily well organized, since city policy is to only provide "tolerances" for vendors who have joined a preexisting organization or formed a new one. Thus, the leaders of the diverse organizations, some with up to 7,000 members, can compel their members to pay high fees and to attend political rallies by threatening to suspend or cancel their membership, which would effectively prevent them from selling on the streets at all. These practices have been observed on numerous occasions by the author and are a constant source of comments and complaints by vendors themselves.(4)

Even if the relocation in the Historical Center were completely successful, it would still leave about 180 thousand vendors on the streets of the Federal District between "fixed stalls", "semi-fixed stalls", "toreros", and "tianguis" of the colonias and comercial zones in the rest of the city.(5) If the DDF wants to eliminate street vending in all its territory, as seemed to be the case according to my own personal interviews with high-level officials of the DDF during 1991-1993 and as Lic. Albores Guillen, Director of COABASTO, commented to the Wall Street Journal in 1993 (November 16, 1993, p A18), it will have to confront this greater problem.

But after two years, even the Historical Center project appears to be failing miserably. In a survey carried out of the new "commercial plazas" in the summer of 1995 it was found that, of the 24 plazas so-far constructed, most suffered from high rates of absenteeism among vendors who, in the vast majority of cases, had returned to selling in the streets, either in the Historical Center or in other areas--sometimes just to be able to pay the high mortgage payments on their new stalls. Only two of the new plazas were "successful" in the sense that most vendors were able to pay the expenses of their stalls out of their stall's revenue. Several markets had over 80% vacancy rates and several of the leaders had told their vendors to cease making payments and had arranged schedules for their members to sell on neighboring streets.

As a result, a true political crisis appeared to be in the offing. Pressured by their members to take a hard line, leaders attacked the government savagely for failing to live up to their bargain, and demanded the right to reinvade the streets pending improvements in the market program, which they criticize for providing poor construction, worse locations and non-existent publicity. Leaders previously loyal to the PRI suggested consistently in interviews that opposition parties were able to offer better deals for their members, and several veiled threats against the system were raised--including the possibility of an alliance between disgruntled vendors and the RUTA-100 union, associated with the leftist "Movimiento Proletaria Independiente" and the EZLN, battling against the city's attack on their jobs.

The economic failure and political repercussions of the plaza construction project are partly based on inadequate planning and the power of street vendor unions themselves in the area. But they are also based on the history of relations between the state and street vendors itself--a history in which, as I will argue in this paper, the political and administrative systems have conspired to create and maintain the powerful street vendor unions in an attempt to coopt the informal commercial sector as a source of political support for the regime since the 1950s.

This article will analyze and compare the two periods of market construction in Mexico City, the first during the regency of Ernesto P. Uruchurtu, and the second the "commercial plaza" project initiated by Ma¤uel Camacho Solis in 1991. I will show the limitations of the state's policy towards street vendors. The central questions in regards to the first period are the following: How did state officials manage to relocate street vendors into market buildings? and, Why did these policies fail in the long-term? These questions are important not only because they illuminate the problems that city officials faced in this earlier period, but also because, as this paper will show in its analysis of the second period, the difficulty of controlling street vending today stems to a large degree from the very mechanisms which the state used against street vending during the Uruchurtu period.

Many have noted the historical parallel between the present market construction program and that of the government of Ernesto P. Uruchurtu, the so-called "Regent of Iron" who dominated the Federal District during an unparalleled 14 years between 1952 and 1966. Appointed to the post of Regent of the Federal District by three presidents in a row, Uruchurtu left his mark by, among other things, building 150 market buildings to house almost 55,000 street vendors during his regency. Uruchurtu's methods in "modernizing" Mexico City were frequently brutal, involving the harassment, illegal imprisonment and beating of street vendors and "paracaidistas", land invaders who's illegal occupation of land was spurred by the strict limits that Uruchurtu put on the growth of the urban area of the city despite its rapidly growing population (Cornelius 1977). Indeed, Uruchurtu was finally chased out of office after ordering the razing of a "invasion" community, suggesting the high level of conflict between his administration and the "informal" subsistence activities of the poor.

Still, while his authoritarian actions towards the poorest members of society eventually led to his downfall, he was still seen as an ideal political figure by the middle and upper classes for whom the order imposed by Uruchurtu, however brutally, had its own merit. As the Excelsior editorialized after Uruchurtu's resignation was demanded by a Congress stoked by public opinion:

"We remember that before Uruchurtu we metropolitans came to believe that the ever-present dirtiness in our markets and the quality of "z¢cos" (open air markets) that were in our most central streets were just part of our national idiosyncracies..." (Excelsior 9/15/66)

It is this image of Uruchurtu that today's political leadership wish to be compared with--the Uruchurtu who cleared the city's sidewalks of vendors. But the failure of Uruchurtu's legacy is evident in the fact that, while the markets he built still stand, the city's streets are again filled with vendors. Is this fact simply the result of later bureaucratic incompetence, or did Uruchurtu's policies themselves help create the basis for the powerful vendor organizations that make the control of street vending so difficult today?

But the law was never implemented as such. Upon entering office in December of 1952, Uruchurtu simply outlawed street vending by administrative fiat. Within a week he ordered the city center cleared of its 2,100 vendors, who were required to relocate into a recently constructed but empty market building, despite the fact that the 1,000 stall market was far from sufficient to handle all the displaced vendors. Still, as newspaper editorials trumpeted, the new Regent appeared to accomplish in a few days what previous administrations had been claiming for years could not be done--he had cleared the downtown area of vendors, and before his first month of office was over he extended the ban to the rest of the city as well. His authoritarian style gained the new Regent the reputation of a decisive leader--and earned him the moniker of "Regent of Iron". The press acclaimed him profusely--an editorial cartoon in the Excelsior showed him as an American football player rushing a football marked D.F. (Federal District) irrepressibly towards the end zone, while the caption read, "May he never stop!".

But the decision to abolish street vending had been made without any negotiation or planning, and the resultant backlash forced the city to modify its plans subject to the availability of suitable alternatives. 2,000 of those affected by the city-wide ban staged an occupation of the city offices and the Director of the Market Office, Gonzalo Pe¤a Manterola, was compelled to personally lead a search for "alternative" sites for the displaced vendors (Excelsior 12/18/52). In fact, the city was forced to allow street vending in most areas until markets could be built to accommodate the vendors in adequate facilities.

Like the government today, Uruchurtu made use of a number of rhetorical claims about the "evils" of street vending to justify his strong action against them in the face of recent legislation that basically legalized the activity. Today street vendors are accused of everything from causing "disloyal" competition for "legitimate" stores, avoiding taxes, to causing air pollution and a general public health threat. Foremost among the rhetoric used against street vendors by Uruchurtu was that they caused traffic congestion and were the cause of inflation because of the purported effect of "intermediarism".(6)

Rather then simply forcing street vendors to give up their trade and replacing them with modern supermarkets, which would have dealt efficiently with the perceived problem of "intermediarism", Uruchurtu began an ambitious program for the construction of covered market buildings to house the vendors who were displaced by the above orders. The markets were built from public funds and rented to the vendors at a symbolic cost, officially in order to reduce the cost to the final consumer but, more importantly, to "bribe" vendors into moving off the streets without resistance. As an official who entered public service at the time explained, in order to entice the vendors into the markets Uruchurtu gave them refrigerators, scales, "...and even then he gave them the maintenance and care of their merchandise with guards, the light, the water..." and many of their other needs included in their nominal rents.

In the market area of La Merc‚d, at that time the largest outdoor market area of the city, market space was constructed for 6,727 vendors. In the area known as La Lagunilla, 2,036 vendors were accommodated and in the commercial area of Tepito 4,488 vendors were placed in market structures. All of these were inaugurated in 1957, although in some cases previous market structures for far fewer vendors had existed before and been replaced. Altogether, in 1957 alone 18,414 vendors were relocated into 36 markets, and between 1953 and 1966 a total 174 markets were constructed or reconstructed for 52,070 vendors raising the number of markets in the Federal District from 44 to over 200. This level of market construction has never been repeated, and the markets built during this period accounted for 77.6% of the 67,066 market stalls in the city in 1993, 26 years after Uruchurtu left office.(7) Even these figures underestimate the contribution of Uruchurtu to the present-day public market system in Mexico City since over a dozen markets were in the process of final planning or construction when Uruchurtu was removed from office.

But if the markets, with their refrigeration units, childcare centers, security, lighting, running water and other "modern conveniences" were the carrot designed to entice vendors off the street, Uruchurtu also wielded a big stick--street vendors who refused or were unable to enter the markets were severely repressed. Guillermina Rico, who was too young to qualify for a stall in La Merc‚d when the market was constructed there in 1957, explained what happened as she stubbornly continued to sell on the street to help support her family:

"Then those of us who were left out stayed to suffer: 36 hours, 15 days, draggings, kickings and beatings... They threw us in jail for 36 hours, and the children were sent to the orphanage... Many of the (street) merchants didn't know prison, but they got to know it for being merchants, for earning their living honestly."

Marta , a vendor who started selling during the last years of Uruchurtu's reign, comments:

"... there was no permission to sell, so since the (street inspector) trucks arrived, we had to find a way of working early. We worked from 5 to 7:30 in the morning because the trucks arrived and removed us. At first they just took the people, and the things were left thrown on the ground. But later, since the government saw that some kept selling, they changed their system again. Then they came back and they took all the merchandise. They let us go free, but they took your merchandise. Everything, they took everything, whether it was money or the merchandise, everything there was..."

A high level of repression was a crucial aspect of the market construction program. Since commercial patterns had become established in the streets, vendors would not enter the markets unless the city could guarantee that their old sales areas would not be taken over by new vendors who could steal their clients. Once they were given this guarantee, the associations of street vendors agreed to enter the markets in mass, and many markets were designed and constructed for a single association of vendors. Thus, vendors who decided to stay in the street were left without any organizational representation to resist the repression against them. All they could do was to attempt to avoid the repression, as these women did.(9)

Uruchurtu was a close friend of President Miguel Alem n with whom he went through college and served as campaign manager and confidant. In 1952, he was appointed to head the Federal District by President Ruiz Cortines in a move that is largely seen as a compromise between the outgoing and incoming presidents to attempt to maintain some powerful positions within the cabinet for Alemanistas (supporters of Miguel Alem n). There is less agreement as to why he was reappointed by Adolfo Lopez Mateos 6 years later. One argument is that it represented continued power on the part of Miguel Alem n. An alternative explanation would simply be that Uruchurtu--who himself was a pre-candidate for the presidency--had become so powerful that Lopez Mateos couldn't get rid of him. (10)

Lopez Mateos was selected as a compromise candidate over Uruchurtu because the latter was too identified with Alem n's supporters within the government, and thus was rejected by the supporters of C rdenas, the powerful left-wing president of the 1930s. But he was a weak president who was dominated by the Alamanistas and Cardenistas in his cabinet. Uruchurtu, on the other hand, had a long history in administrative positions in the PRI and in state and national government, and had a powerful cadre of allies.

In 1964 Uruchurtu was again a pre-candidate for the presidency on behalf of the PRI. However, Gustavo Diaz Ordaz was elected and many argue that tensions between the two figures were apparent from the beginning, although Uruchurtu was again reappointed to the position of head of the DDF. While Uruchurtu's conservative stance was similar to that of Diaz Ordaz, his power as Regent and behind the scenes became a threat to the President's authority. As one scholarly work in its discussion of the conflict between these two political figures notes:

"A President generally guards the expectation that his collaborators will not exceed him in power, so that he can shine (destacar) personally." (Lerner de Sheinbaum & Ralsky de Cimet 1976: 395)

It is highly significant, however, that the final downfall of Uruchurtu was precipitated by his continuing policies against informal-economy linked movements within the capital city. Besides his attacks on street vending, Uruchurtu was also known for his harsh repression of land invaders, who thwarted his attempt to limit urban growth and imposed high costs on the city for the purchase of land and provision of urban services in areas that were often only marginally habitable or virtually unserviceable due to topographic features or soil conditions. But his practice of ordering such communities to be razed was an unpopular one which provoked many ideological problems for the PRI. When the city bulldozed a community of 400 land invaders during a downpour four days before the "grito" of independence (September 15, 1966), members of congress lined up to denounce the act and the PRI skillfully distanced Uruchurtu from the President before censuring him and forcing his resignation. During the congressional debate, land invaders and street vendors were both held up as noble victims of the cruelty of Uruchurtu's policies. One columnist argued that the attacks on street vendors had formed part of the Uruchurtu's "dictatorial pattern" against "...the poor indigenous women who in the use of their inalienable right of survival sold their miserable merchandise on the urban sidewalk." (Carlos Alvarez Acevedo, Excelsior 9/15/66:7A) Four powerful worker unions affiliated with the PRI, including the CTM and the Federation of Revolutionary Workers (FOR), also attacked the "...injustice that is committed against the humble merchants, ambulatory vendors, who are imprisoned and whose goods are confiscated." (Excelsior 9/15/66:1A)

The paradox here is that Uruchurtu's policies towards street vending--the construction of markets and the resultant necessary repression of vendors--was originally a key component of his power base. Not only did the policy please the middle classes and provide a substantial source of lucrative contracts for the Mexican construction industry, it also enabled Uruchurtu to organize and manipulate street vendors politically for his own purposes.

Administratively, the civil associations enabled city officials and market administrators to negotiate with a single body of vendors that represented--or at least could be held accountable for--all the vendors in a given street or market. "You cannot negotiate with 100,000 vendors--it is easier to negotiate with a few leaders," as several officials have told me. Politically, the civil associations were also obliged to affiliate with the PRI and support political actions on behalf of that party. More importantly, Uruchurtu used the associations to advance his own career. The skillful manipulation of the market construction program was already paying off powerfully in the political realm by 1958. On February 12 of that year a rally of 40,000 market stall-holders and their families showed up for a huge political rally in support of Lopez Mateos, and the head of the PRI in the federal district, Rodolfo Gonzalez Guevara, noted that the 1958 campaign was the first in which small merchants had played a substantial role--something he attributed directly to the market construction program. More significant was the fact that leaders of the market vendors overtly connected their support for Lopez Mateos with his promise to continue Uruchurtu's policies within the city. (Excelsior, 2/1/58 and 2/15/58)

The timing of the inauguration of markets during the two presidential terms in which Uruchurtu served as the Regent of the Federal District further indicates that the inaugurations were calculated to garner immediate good will and support for himself and the PRI. By comparing the number of vendors who received a stall during each two-year section of the sexenios of Ruiz Cortines and Lopez Mateos, when Uruchurtu was in full charge of the Federal District, it becomes clear that the last two-year period of each sexenio--the period directly before the presidential elections--were the object of a much higher than normal level of market construction.

It was in fact this very politicization of the market program and of street vending that ultimately spelled the doom of the program itself. While the markets were used to entice vendors into the PRI, this practice was self-limited by four factors. These will be mentioned here, and will be discussed in depth in the next section. First, while the policy of market construction had created a large pool of associations loyal to the PRI, this was obtained at a huge financial cost. Secondly, by 1966 the vast majority of street vendors were either in market structures, or had rejected them in favor of a return to the street. Thirdly, there was a tendency for support from market vendors to decline over time. Finally, the enormous subsidy of market vendors led to profiteering by vendors and actually led consumers to request the return of street markets.

THE COST OF THE MARKETS

THE REJECTION OF MARKETS

Constructed on available lots or areas of cheap land values, the markets were usually not as centrally located as the street markets they replaced, meaning that fewer clients came to them. While new clienteles were built up over time, many vendors seem to have left during this initial phase.(11) In addition, the symbolic rents were by no means the only cost of entering the markets: besides the continuation of corruption (many of the administrators were ex-police officers, and highly susceptible to such activity (Eckstein 1977)) city officials also used the markets to gain control over the commercial activities of the vendors in ways they found difficult or impossible to control in the street, such as imposing regular hours and regulating "giros" or product lines. A third factor was that the simple change of locale from the public thoroughfare to a market changed the nature of selling.(12) Again, this required vendors to change their marketing strategy, an adaptation that few were prepared for.

For these and other reasons, the relocation of thriving street markets into closed markets damaged the commercial allure of whole neighborhoods. By mid-1953, even the established merchants of swanky Polanco were urging the city to let street vendors come back, because they claimed their sales had dropped by 50% while other neighborhoods where vendors had remained experienced an increase in sales ("Distrito Federal" in Excelsior, 6/20/53). Many vendors from this period whom I interviewed claimed they suffered even greater losses and their savings were depleted. Some took to selling door-to-door, some had to get factory jobs, while others simply went to areas where markets were under construction to take advantage of the temporary permission to sell in the street (and in some cases to amass stalls for resale or later use).

A good example is presented by the working class area of Tepito which, located near the downtown area and one kilometer from the Z¢calo, had grown into a major street market area since late in the last century (V zquez Torres 1991). Used goods of frequently dubious legal origin--giving the area the name of the "thieves market"--were sold from wooden shacks constructed in the middle of the street that in many cases also doubled as homes. These vendors presented few protests when the markets were constructed and their stalls and homes were destroyed in 1957, but Alfonso Hern ndez Hern ndez, a Tepito historian and an official in the Cuauhtemoc Delegation, maintains that after three years vendors began leaving the markets, faced with the prospect of complete financial ruin if they stayed in them. In some cases they sold their rights to their stalls,(13) in others they just left the stalls vacant--giving up all right to them. One of the Tepito markets--which now successfully bills itself as the "worlds largest shoe market" after each of its 700 stalls began specializing in shoes--is now reportedly controlled by seven families who bought out most other vendors as they left the market.(14)

THE DECLINE IN SUPPORT FROM MARKET VENDORS

SUBSIDIES AND PROFITEERING

The above factors reinforced each other to produce a steady reduction of the political rate of return from the market construction program over time. While the construction and administrative costs mounted, the political benefits dwindled until the fantastic expenditures no longer made sense. At the same time, the market construction program implied a massive enforcement campaign to force street vendors into the markets on the one hand and to prevent them from competing with the markets on the other. Thus, even while the market construction program produced steadily reduced political benefits in terms of decreasing support from market vendors, it produced more and more political liabilities on the other hand in terms of its attacks on street vendors. In a word, these policies were no longer politically or economically viable. The street vendors that existed were no longer amenable to entering markets, and by enforcing stiff laws against them the PRI only succeeded in alienating them while the city administration--and by extension the national administration--merely looked authoritarian and dictatorial on the national and international level.

One must also consider the broader political aspects of the Mexican political scene. A decade and a half of growth had woken hopes of improvement in both political and economic aspects of the Mexican domestic scene--neither of which were nearly as fast coming as had been anticipated. At the same time, political control had been virtually monopolized by the right-wing of the PRI with a succession of anti-populist presidents. Growing political unrest began during this period, particularly among the middle-classes. The student movement--which ultimately was repressed by the army in a massacre costing hundreds of lives in July of 1968--already had a strong and growing presence. As Cornelius notes (1975), it was simply not possible for the state to continue alienating a growing proportion of the "popular" classes during this period. Just as the PRI became intimately involved in land invasion issues during this period, so it also needed to find allies, not enemies, among the street vendors.

Thus, Uruchurtu's market construction program had changed from being a political asset to the PRI in 1958 to a financial and political liability of huge proportions by 1966. The razing of a community of land invaders was not simply an excuse to get rid of a political rival, but a symptom of the very danger that Uruchurtu posed for the regime. He now got in the way of the PRI's attempts to cement support among the lower classes because of his refusal to allow land invaders or street vendors to become incorporated within the political-administrative structure.

But the market construction program, as it was implemented by Uruchurtu, established two precedents that are crucial in explaining the power of street vendors today. First, by requiring street vendors to be members of a civil association in order to be considered for a market stall or to be "tolerated" in the street, Uruchurtu laid the groundwork for today's powerful street vendor unions. Instead of the "voluntary" civil associations envisioned by the 1951 market regulation, Uruchurtu's policies gave complete power to association leaders over their members, since it gave them, rather than the city, the ultimate power over the livelihood of the individual vendor, who could not get a stall or a "tolerance" without the leader's blessing.

Secondly, Uruchurtu's policies not only firmly centered the market vendors within the PRI, they also provided a model which future associations could follow of exchanging loyalty to a political patron (usually within the PRI, but later also within other parties) for reciprocal support of their "right" to a market or a street. Above all, street vendors learned that if they could hold onto an area for long enough, the city would recognize their "right" to it, or, the next best thing, construct a market for them nearby.

Yet within a year of these statements, these same officials unveiled a program designed to eliminate street vending from the map of Mexico City, or at least to begin to do so, in an ambitious program of market building construction that would herald the formalization of this vast economic sector. This change in the official posture towards street vending closely followed the 1991 mid-term elections in Mexico, in which the PRI made a surprising come-back. After almost losing the presidency and control of congress in 1988, the PRI swept the 1991 elections with only a few charges of electoral fraud. Street vendors, who traditionally had been used as "acarreados" (paid supporters) by important groups within the PRI to "legitimize" elections by a show of "supporters", now perhaps didn't seem to be as necessary to a regime focusing its sights on the "modernization" of Mexico City in anticipation of the North American Free Trade Agreement. Street vendors were "hold overs" of the traditional past, and the Salinas regime wanted to project a different "image".

Although the local Assembly for the Federal District (ARDF) had spent two years writing a new ordinance to regulate street vending, this was shelved after the elections while the city began to plan the removal of ambulatory vendors in the city's Metro rail system as a prelude to removing vendors from the Historical Center and, later, the rest of the city. A series of "negotiations" between officials and street vendor leaders began in which city officials urged the leaders to sign a vaguely worded general statement that promised to "reorder" street vending in the central area. Later it became clear that these terms were taken by the city to mean the relocation of vendors into specially built market buildings.(15)

The first step was to clear the city's metro system of vendors. A month after the 1991 elections officials began to talk of an upcoming crackdown, and the Metro carried out a propaganda blitz identifying the vendors with safety hazards. The actual crackdown came six months later in January, 1992 after a fight between a vendor and a passenger resulted in a fatal shooting. Using the shooting as an excuse, the city clamped down with the use of hundreds of riot police. The metro operation was significant in two regards, despite the fact that it was never 100% successful. First, it showed vendors that the city could use force if necessary, and thus pressured vendor leaders to participate in the negotiations. Secondly, the shooting provided substantial public support for the city's actions, suggesting that public opinion was moving against street vending and in favor of the city's new law-and-order policies.

THE HISTORICAL CENTER: NEGOTIATION AND CONFLICT

But also present in the Historical Center were several associations of small merchants who claimed to represent the many small shops that do business in the area. While a certain level of competition had always existed between these merchants and the street vendors (although to a certain degree their interests were also complementary since the street market attracted huge numbers of clients to the area), the conflict between them intensified after the passage of new fiscal regulations which compelled small merchants to purchase "fiscal cash registers" that the tax authorities hoped would stem the high degree of tax evasion among shopkeepers. Despite the fact that many small shopkeepers simply never used the machines,(16) they still had to purchase them while their "informal" street colleagues avoided both the taxes and the machines. The reaction to this measure was harnessed by some of the small-business associations to increase their own membership roles and buttress their importance in the area by laying the blame squarely on the street vendors.

The most militant of these groups, "Proc‚ntrico", was formed by a merchant named Guillermo Gazal Jafif to push for the elimination of street vending in the city center. Originally ignored by city officials and snubbed by the city's official Chamber of Commerce (C mara Nacional de Comercio, or CANACO), this group became highly successful in gaining national attention for their cause and pushing the city to carry through on the "reordering" of street vending.

During this initial period, several street vendor organizations began extending their areas and augmenting their numbers of vendors--to guarantee a larger number of market spaces, or at least a better negotiation position. Others took advantage of the plans by charging their vendors huge fees to "guarantee" their places within the future markets. Officials added to the confusion by claiming that each organization had to create a fund to pay for the down-payment of their project--but vendors were still charged individual down-payments by the city after paying into funds supposedly maintained by the associations for the same purpose.

What made the policy more difficult to implement was that, in keeping with Salinas' neo-liberal policies of eliminating subsidies, vendors had to pay for the full cost of the stalls, breaking sharply with Uruchurtu's policy in which markets had been constructed by the city and then rented to vendors at a subsidized price. Possibly, the goal was to prevent the new market program from collapsing due to lack of financing, as Uruchurtu's program had done. As a result, the Salinas government planned to build "markets in condominium": "commercial plazas" in which stalls would be sold outright to individual vendors and an association of "vendor-owners" would own and maintain the building infrastructure. But when a "test market" was constructed in the nearby "La Merc‚d" district in 1991 (The "San Ciprian" market for 1,100 vendors), only 20% of the stalls were initially occupied despite the fact that the street vendor association had signed an agreement to occupy it and that the vendors had already paid the down payments. The failure of the market was acutely embarrassing for the city given its need to convince the Historical Center leaders to follow the same model.(17)

The slow pace of action finally brought a response from the leader of Proc‚ntrico, who had attacked city officials as early as January of 1989 for allowing the street vendors in the center to "steal" 10,000 million pesos (U$4.5 million) in sales from the small stores. In January of 1990, after obtaining a supreme court ruling against street vending, he threatened that his members would stage store closings to enforce the ruling, leading the Regent to accuse him of "bordering on fascism". In late August and early September of 1991, Gazal again appeared in the public light, urging that the vendors at least be put into bazaars.

RED MONDAY

This time the vendors staged a counter-demonstration against Gazal. The crowd of enraged vendors marched up to where he was standing and, after a short shouting match, bright red tomatoes started falling out of the sky besmirching Gazal's cashmere suit. After a few seconds, Gazal and two assistants who were with him began to run, and the crowd ran after them. The following notes were taken while watching a video tape held by the vendors' association:

It shows young people running to the side of and in front of Gazal throwing tomatoes at him and shouting "get lost" and less pleasant things. But even though they could have cut him off easily there is no attempt to do more than throwing tomatoes. No stones are seen. When Gazal gets to the corner of Brazil and ran around it with his assistants, several of the attackers follow him, but are called back by the others. The film shows Gazal and his assistants getting into two sports cars parked on the street and taking off as fast as they can. "Get lost, damned ______," a voice is heard. Anti-semitic profanities were also heard frequently on the tape.

But the event was clearly welcomed by Gazal who, in the opinion of many members of the press, had been hoping to provoke such a response. One noted that even several hours afterwards, Gazal was still in his tomato-stained suit waiting for a television crew to interview him. He was first-page news the next day, and was carried by the top evening news program in the country the same evening, sparking massive criticism of the protestors and of street vendors in general, and helping to satanize the Historical Center vendors in the same way that the Metro shooting had stripped the Metro vendors of any public support.

However, in contrast to the unanimous press opposition to the metro vendors when they were attacked, several key newspaper sources supported the street vendors this time, attacking Gazal for being an opportunist who simply wanted to get attention in order to win a power struggle with the CANACO. A political columnist for La Prensa, a popular newspaper that generally attacked street vendor organizations, went so far as to accuse Gazal of committing fraud against investors who were trying to build markets in the city, and the same columnist took delight in noting that even after the above events, Gazal couldn't get an appointment with the Secretary of Governance.

But Gazal was in fact highly successful. Not only was vending on the street involved suspended immediately, with a 24-hour guard of over 60 riot police, but a week after the tomato attacks a shuffling of major posts and responsibilities within the administration was forced on the city by President Salinas. The Secretary of Governance who had ignored Gazal was reassigned to a much less important position. With him, following Mexican administrative tradition, went his entire staff, including those who had begun to take the office beyond its purely oversight role into the planning and implementation of the relocation of street vendors in the Historical Center.

At the same time a new "Coordination" was formed to take over the functions that the Secretariat of Governance had begun to perform in coordinating the market construction with a specific mandate of coordinating the construction of facilities for the relocation of all street vendors in the Historical Center. The Secretariat became a purely administrative office once again, with no direct policy implementation responsibilities.

Gazal was not the only factor influencing this process. Interagency rivalries were also influential--with the "red monday" event providing an excuse for attacking the Secretariat of Governance by a number of different groups that were in danger of having their own power undermined by the growth in the Secretariat. Not least of these were vendors themselves, as well as the PRI. But within the administration the biggest enemy would seem to be the 16 administrative sub-units of the city. Called "Delegations", they were each responsible for administrative and regulatory functions within a specific territory--functions which the secretariat had begun to take over. While the officials I spoke to in the Secretariat of Governance denied that any tensions existed between their office and the delegations, the fact that they were writing a budget that included over 80 new personnel whose sole purpose was to make sure that the delegations were doing their jobs properly indicated that relations between the secretariat and the delegations were far from tension-free.

The formation of a separate agency (later merged with COABASTO) to coordinate and negotiate the construction of markets, also supports this explanation. By removing the implementation of the street vendor policy from the Secretariat's office the city appeared to be both allowing for centralization as well as keeping the secretariat's office from becoming too powerful. Significantly, the new agency had no direct administrative powers, and had to rely upon the delegations to actually implement policy decisions.

A third factor also appears to be at work. Cutting back on the authority of the Secretariat of Governance also satisfied the concerns of leaders, who were afraid that the secretariat's office would have no limit on its power and mandate if it became directly involved in the affair. One of the immediate results of the creation of the new office was that--while the secretariat was stripped of any direct responsibility for street vending, the new agency was only given a mandate to begin construction of markets for street vendors from the Historical Center, when the secretariat had been discussing the elimination of street vending from the entire city. The result was a far more limited mandate with less powers for carrying it out on the part of the agency entrusted with eliminating street vending.

Backed by presidential authority and spurred by organized groups in civil society, the plan advanced slowly but surely over the following two years. A special credit institution, the Fondo de Desarollo Econ¢mico y Social del Distrito Federal (FONDEDF) was opened as a subsidiary of Nacional Financiero, S.A., a government-owned bank, to manage the financial side of the program, putting up 43 per cent of the 580 million new pesos that the project required. (Approximately 175 million US dollars at 1994 exchange rates) while most of the rest was provided by the Banco Nacional de Comercio Interior, another government-owned financial institution.

Since street vendors generally did not have proof of their income nor were they able or willing to put up other property for security on the debt they were to incur by entering the markets--generally from US$5-10,000 depending upon the size and type of stall, of which 10% was to be paid as a downpayment, with monthly payments stretched over 6 years, the city had to function as the guarantor of all the financial credits provided for the market program. Thus, even though the city was operating in a new "neo-liberal" mode very different from the 1960s market program, it still incurred a huge financial risk as a direct result of the pressures noted above (DDF-COABASTO, 1994).(18)

Another important clue is given by the many statements by officials who argued that street vending was incompatible with the modern image that Mexico City was attempting to project in light of the North American Free Trade Agreement with the United States and Canada. While street vending had been largely protected during the period of economic crisis in the 1980s by the argument that it provided economic opportunities for the unemployed, interviews with officials after the 1991 elections showed a pervasive attitude that vendors had become wealthy and could now begin to pay their own way. On the other hand, a misplaced belief in the ability of NAFTA to reinvigorate the economy and provide new jobs also provided a rational for eliminating the opportunities that street vending provided.

At the same time, mounting pressure from small-business associations to eliminate the "disloyal competition" of street vending was itself largely caused by the increasing formalization of that sector, causing a greater distance between small businesses and street vendors in terms of the formal costs they faced.

Finally, the 1991 elections seemed to show that the PRI could win elections without resorting to massive electoral fraud, and thus the role of street vendors as "acarreados" became less important to the state as a whole. While many street vendor organizations still maintained strong connections to the PRI and were still politically important, as will be argued below, this provided an "opening" for city officials to take action against street vending.

But the limitations of the political system were still apparent in the relocation process itself. Most importantly, it was apparent in the ability of street vendor organizations to strengthen themselves even as they acquiesced to the relocation program itself.

On the surface, the project appeared to be a complete success. The Christmas Season of 1993-94 was the first season in recent memory in which the number of vendors in the Historical Center did not double, and by the summer of 1994 the streets of the area seemed curiously empty. Traffic and pedestrians moved unencumbered by the thousands of vendors who had crowded the streets and sidewalks. While newspaper stalls had not been affected by the "Bando",(19) and a special class of candy stalls seemed to be also inexplicably immune to the ruling, the number of vendor stalls had dropped to one or two per block as opposed to the hundreds that had existed previously.

Furthermore, the city was implementing plans to begin to fully "formalize" the vendors by making them subject to the tax code as small businesses. Those who only a year before had been informal street vendors were now located in new market buildings, had been counted by the commercial census, held mortgages, and were to begin paying taxes.

But underneath this surface many cracks rapidly appeared. While a few of the markets that were exceptionally well placed or planned (usually those right next to Metro facilities) had a thriving business, many of the markets were barely surviving. Vendors opened their stalls, but reported low sales that did not cover their monthly payments. And in a brief survey of markets on a weekend, when sales are traditionally high, at least half the stalls appeared to be permanently closed. The worst area was La Merc‚d, a market area where over 3,000 stalls had been built in the 1950s and where an equal number had been constructed to accommodate a new generation of street vendors in 1993. While the old market buildings were doing the same level of business as before, the new markets were almost entirely empty. One market built for 1,500 vendors had only 20% of its stalls open on even the busiest days, mostly on the side of the building closest to the old market. Despite the fact that the stalls had only just gone into repayment, a large number of stalls had already been repossessed by the FONDEDF due to lack of payment on their mortgages as their owners gave up on them and forfeited their downpayments.

By the summer of 1995, street vending had returned a large numbers on the streets, with an estimated 3,500 vendors in the area and larger concentrations around the perimeters of the prohibited zone. Many vendors reported that their leaders gave them permission to sell in nearby streets as long as they also kept their stalls open, apparently with tacit approval of the authorities. The reoccupation of the streets followed a double logic: not only was it necessary on immediate economic grounds(20), it was also necessary to prevent other groups or new groups of vendors from taking over "their" streets. The failure of the city to enforce the "bando" was thus blamed by leaders for "forcing" them to reoccupy the streets in violation of the very same regulation.

Even in San Ciprian, which by 1995 had been open for five years, most stalls were still unoccupied, and those vendors who did spend long hours waiting for the occasional customer to drift through complained about the difficulty of making their mortgage payments while at the same time restocking their merchandise, remodeling their stalls to attract more clients and eking out enough to eat from the low turnover they experienced. Again, many had to take second jobs, or close their stalls while they paid it off and saved enough money to remodel and stock up. That vendors were willing to do so showed that they thought of the stalls as investments that would hopefully pay off over the long term, and their willingness to work double shifts (or triple shifts in the case of many of the vendors who were women) and to forgo present income is a credit to their entrepreneurial spirit. But the vast numbers of empty stalls, stalls under repossession and stalls that were being sold illustrate that many of the vendors simply could not pay these high prices, and would be forced to look for employment elsewhere, or back to the street.

The resale of stalls in itself posed a difficult problem. While the stalls were in repayment, resale or even lease of the stall was ruled out by the mortgage contract. But the number of signs posted by the FONDEDF to remind vendors that this practice was illegal seemed to suggest that it was fairly common. Nevertheless, once a stall was paid off, there was no limitation on its disposition. Unlike the public markets constructed previously, in which the stalls were technically inalienable from their original beneficiaries (but in which corruption usually made their resale and lease fairly straightforward), in the new markets someone with sufficient capital could quite legally buy up all the stalls in any given market, using various forms of pressure to push the original small vendors out of the way.(21)

During this period the associations of street vendors had been far from defeated. On the contrary, the acknowledged policy of the city was to work in conjunction with the street vendor leaders as much as possible. As in the earlier period, most of the markets were designed for specific associations of vendors, and the leaders were given a number of significant powers in the process. First, city officials negotiated directly with the leaders in the Historical Center as a group and as individuals in designing and setting the rules for the markets. But more importantly, only vendors who were accredited members of an association were allowed to purchase space in the market program. Thus, since leaders defined membership, leaders were made the "gatekeepers" of the market program, with the power to control access to market stalls and to distribute the best stalls to their friends or those who paid the most for them. Thirdly, it gave the leaders control over the markets themselves which opened up the possibility of cacique-type abuse of their powers in detriment of the small vendors who were supposed to be the beneficiaries of the program. Finally, it meant that leaders, far from being weakened by the market construction program that was supposed to destroy the associations as it made individuals into property owners, were made more powerful as they gained control of the markets and continued to have the capacity to threaten to lead a reinvasion of the streets if they deemed it convenient. As one official noted, "One of the biggest problems was that of giving the distribution of stalls to the leaders."

Given the fear among officials that the associations would be able to successfully stall or fight the market program by simply refusing to leave the streets, the policy to leave substantial power in the hands of the associations was a rational one. But it meant that the associations maintained strict control of the markets in most cases, with the leaders becoming in many cases the administrators of the markets themselves. Each market was legally required to form an executive committee to take care of maintenance and communal problems but, given their control over the vendors, the committees were in almost every case controlled by the associations that the markets were built for, and the leaders continued to charge the same fees they had charged on the streets. Some associations extended their control further by using their own reserves (built up by vendors' fees over decades) to lend money to vendors for the downpayment, or by acting as the guarantors on the mortgages, thus getting control over them as individuals.(22) At the same time, while officials recognized that the leaders "continue to benefit because they are left in control of the executive committee", official policy is that only the vendors themselves can change this situation, since intervention by the city into the internal affairs of the markets or associations would be seen as "authoritarian and undemocratic", (despite the fact that the city's policies gave control of the markets to the associations).

But by not attacking the associations, the greatest fears of the administration could still be realized because, although the leaders had signed the agreement in the Historical Center, they constantly threaten that if the new markets did not work to their satisfaction, they would reinvade the streets in the area. Indeed, the pressure to do so built up to a crescendo during the weeks before the August 1994 Presidential elections and again in the summer of 1995. One of the most important leaders in the area, herself a candidate to the ARDF on the part of the PRI, savagely attacked the city for not living up to its end of the agreement and argued that this gave her association a legitimate right to go back on it.

In July of 1995, the legal advisor to another powerful leader affiliated with the PRI noted that, "we can give the system 80% of the seats (in upcoming elections for "Ciudadanos Consejeros"), but they don't seem to realize how important we are," and complained about favoritism shown to opposition party groups. "If we became members of the opposition, we would probably be better off," he noted darkly.

A leader of a small association noted in 1994 that, while he was loyal to the PRI, "People are thinking ... if after the elections they don't offer us a solution for the markets they are going to go back to the streets--whatever opportunity they see they'll take." But they had to wait for a large association to take the first step: "If I had the number of people they have I would do it. If I did it now they would squash me with one hand. But I don't know--sometimes its worth it to take a risk," he added. Furthermore, as this leader pointed out, the competitiveness among leaders meant that as soon as one association started to reinvade the streets, the others would be compelled to follow to prevent other groups of vendors from taking over the streets that over the years they have considered as "theirs".

In 1995, another leader went even further, asserting that "Nobody is affiliated with the PRI out of conviction--we are all with the system out of convenience. And there may come a time when it is no longer convenient." While he remained loyal, he noted that militants of the leftist PRD and even the RUTA-100 union, one of the most radical trade union movements in Mexico, had approached them to discuss the possibility of joining forces. "If we were to join the opposition, just imagine what would happen," he noted, adding "we don't want a confrontation because ... it would be unequal... they have 20 or 30 thousand granaderos (riot police), that's why we don't want to get to that level. But if they don't resolve our (problems) we don't know what level we'll get to."

Furthermore, while many leaders and vendors expressed their respect for Camacho Solis, they expressed disgust for the series of officials who followed him after his resignation in late 1993. As one leader complained about the regent, Oscar Espinoza Villareal, and other officials in 1995, he noted, "Camacho was something else--Camacho came to La Merc‚d. This one doesn't even show his face--he lacks political skill... There is no one in the government who can say, 'Lets do this is that'. All they want is 'negotiation'. The Regent doesn't show his face, the delegate sends us to the sub-delegate, and he send us to his secretary. What is she going to fix? Nothing!"

While in the 1994 presidential elections none of the associations that were affiliated with the PRI left the party to join the opposition and many of the leaders and vendors I spoke with expressed continued support for the PRI, by 1995 many more leaders and individual vendors were expressing their discontent. Leaders were under pressure not only because of the problems with the markets, but also because of a rising sentiment among vendors themselves that their role was becoming superfluous. Whether they were in markets or on the streets, many more vendors expressed their independence from their leaders. Many noted that leaders no longer really existed, since they were now owners of their own stall and didn't need the leaders to protect their rights anymore. And even vendors who continued to sell on the street were less likely than before to pay fees to their associations, since they considered that the leaders had given up their claim to the streets when they had signed the agreement with the city.

Furthermore, the financial crisis of January 1995 further affected stall owners by depressing sales even more and raising interest rates on their mortgages. As a result, many vendors found that their debt increased even when they made their monthly payments, and many simply refused to pay them.(23)

Thus, leaders were compelled to demonstrate their "usefulness" by taking a hard line with the city and insisting on the ability to stay in the center's streets. One organization went so far as to demand that the city allow their members to sell in the streets in order to earn enough money to pay off their market stalls. By reasserting their rights to the streets, they also revealed what might have been their strategy from the beginning: the ability to have their "plazas" and the streets, too.

COMPARISON: URUCHURTU AND CAMACHO SOLIS

In both cases, the motives for relocating street vendors into closed markets stemmed from a set of political motives rather than economic ones. Also, the implementation of the policy in each case was also designed to extract the maximum political benefit, and to limit political liabilities. In both cases I have shown that electoral considerations, both for the PRI as well as for Uruchurtu and Camacho Solis, were highly important. While Uruchurtu banned street vending immediately upon entering office, it took 5 years before the first group of vendors were actually relocated into markets--on the eve of the 1958 elections. In the latter case, street vending was only banned after the PRI had apparently regained their position electorally in the 1991 elections. The implementation of the policies--the organization and cooptation of vendor groups in the case of Uruchurtu, and the strategy of giving inordinate powers to the same organizations in the latter case, also cemented corporative support for the system from these organizations. At the same time, the construction process benefitted construction companies and, perhaps most importantly, projected a "can do" image for the middle classes that garnered substantial support in these sectors for their respective presidential bids.

Another similarity between the cases is the lack of initial commercial success of most of the markets in question. While Uruchurtu's markets are today commercially successful, they were initially poorly received by the public and vendors, for the reasons noted above, to the point that an American researcher in 1968 predicted that most of them would fail and should be replaced by supermarkets (Pyle 1968). In the latter case, the failure is all too apparent at present, although many of the new "commercial plazas" may later build up a clientele in the same manner as Uruchurtu's markets did. Much of this failure itself is due to the poor level of planning of the location of the markets and distribution of stalls within them. The emphasis on rapid construction for the maximum number of vendors led to hastily conceived floor plans in which stalls were lined up like cages in a box--all the more so in the case of the recent project, in which all but two plazas were conceived of, designed and built within the two year period between 1991 and 1993.(24)

The most important economic difference between the two cases is the way in which costs were distributed. In the first case, the markets were built by public funds and then rented at a heavily subsidized cost to vendors, partly to secure their compliance and partly to secure their political support. In the latter case, the markets were financed by government-owned banks but sold on credit to the individual vendors, who were obliged to pay a downpayment on receiving their stalls and then monthly payments spread over 5-6 years. On the one hand, this change would tend to prevent the build-up of costs for the city that led to the demise of Uruchurtu's market construction program after 14 years. On the other hand, the immediate costs may mean that the present program will fail to last even that long as vendors default on their loans and push the government banks into fiscal crisis. The assumption officials made--that the vendors earned enough to afford these costs--was never investigated, nor was allowance made for an extended period of lower sales as the new markets were "accredited". As a result, most of the mew "commercial plazas" were in trouble even before the 1995 economic crisis made the situation worse.

But there were also two important administrative differences. Uruchurtu governed the Federal District for 14 years, during which time the market office was directed by a single individual. This gave the period a level of administrative consistency that simply did not exist in the latter period. Simply put, Uruchurtu was around long enough to see his policies through. Camacho Solis, on the other hand, resigned from the Regency before most of the markets in the initial phase (the Historical Center) were finished. While general policy lines remained consistent in the following two years, there were two different regents with completely different staffs and different priorities.

Another administrative difference was the organization of the government itself. In the 1950s and 60s the Federal District was a single political entity firmly in the control of the regent, and only one agency--the Market Office of the Treasury--was responsible for authorizing, planning and regulating markets. But by the 1990s the city was divided into 16 "delegations" and the regulation of the markets and street vendors was divided between offices in the delegations, the "Secretariat of Governance" in the DDF, and COABASTO. The situation in the Historical Center was further confused by the fact that it was divided between two separate delegations. As a result, inter-agency rivalry over responsibility and bureaucratic as well as political (and corruption) benefits in the market program was far more problematic.

But the more important legacy of Uruchurtu is the way that he helped pattern the relations between street vendors and the city that later allowed street vending to re-emerge at unprecedented levels. It was never the intention of the new city officials who replaced Uruchurtu to allow the spectacular growth of street vending that occurred in the following three decades. Rather, they simply wanted to reverse a situation that had become politically untenable for the PRI and economically untenable for the state. However, Uruchurtu's policies had radically politicized street vending by using the market construction program as an incentive for street vendors to organize within the PRI and as a reward for loyalty towards his own political objectives.

After Uruchurtu was forced to resign from office the market construction program was scrapped in favor of new programs, such as the Metro. But officials continued to compel groups of vendors to organize into associations that had to prove their loyalty to the PRI in order to be "tolerated" in the streets. For example, by 1968, already over 10,000 vendors were authorized to sell in "tianguis" (Pyle 1978). Officials hid the fact that they were allowing street vending to re-emerge with the fiction that the groups were being "tolerated" only until markets were built, essentially applying the letter of Uruchurtu's policies but denying the spirit since neither the officials nor the vendors had any interest in the construction of the markets. Still, as noted at the beginning of the article, by compelling street vendors to form or join associations, the city gave the associations the ultimate power over individual vendors, allowing the leaders of the associations to use their members in any way they saw necessary in order to defend their interests--increasing their memberships, increasing their space on the street, and ultimately increasing the "phenomena" of street vending to a level unprecedented in Mexican history.

On the other hand, the Historical Center project appears as a farce--a hollow mockery of Uruchurtu's program. Faced with powerful street vendor unions, officials were obliged to give them substantial power over the relocation program while the rapidity with which the program was put into effect led for the most part to poorly designed and distributed commercial structures with little economic viability. The relative failure of the project is shown by the reemergence of street vending in the limited project area within a few years, compared to the much longer period of decomposition of Uruchurtu's program. Rather than weakening the power of the unions, in fact, the program has led to a potential radicalization of both leaders and rank-and-file members that the use of massive public force will only serve to augment.

This analysis of the history of street vending in Mexico City suggests a number of conclusions about the capability of organization in the informal economy. First, it is clear that organization, initiated by the state to coopt street vendors politically and control them economically, became an effective mechanism by which street vendors could create access to a necessary public good (street space) and to defend it against attempts by the city administration to restrict it or take it away. According to resource mobilization theory, successful social movements are those in which organized interest groups can mobilize key resources in their struggle. In this case, by compelling vendors to organize within the ruling party, the state itself gave the organizations the ability to mobilize their members effectively. The fact that vendors were divided into hundreds of different organizations, normally seen as a strategy of divide-and-conquer, actually weakened the administration's capacity to control them given the competitive dynamics between the organizations: any organization that agreed to proposals for relocation risked losing control of "their" streets, as we saw in the case of the Historical Center project.

Secondly, while compulsory organization often led to the emergence of authoritarian leaders who often ran their organizations as personal businesses(25), they were also required to represent the interests of their members in some respects, but not in others. For example, leaders had little interest in acquiescing to administration demands to restrict the number of vendors they had or to relocate their vendors because this could entail the loss of their members, who would shift their allegiance to other, less docile, leaders. But, when they were forced to submit to the market relocation program they took little interest in the success or failure of their specific plazas, concentrating instead on controlling access to the plazas and attempting to control the plazas themselves. It seems in some cases that leaders actually planned to eventually recontrol street vending in the area anyway, and thus actively undermined the plaza project.

Thus, the system of leadership allowed effective resistance to attempts to remove street vendors from the downtown area of Mexico City, as it has in other areas of the city. In that sense, it has been able to protect the interests of street vendors as street vendors. But it is limited by its own dynamics to the interest base of the leadership itself. Since that interest base is represented by control of the streets, it coincides with the interests of those who wish to remain street vendors. But by the same token it is not effective in representing the interests of those who do wish to make the transition to market vendors. While this presents a specific problem for Mexico City's authorities, it presents another problem for the analysis of urban social movements.

A colleague recently asked me to identify "good, non-exploitative" leaders among street vendors in Mexico City, and this is the problem: Often the best leaders are precisely those who do exploit their members and are authoritarian, because this makes them responsive to their interests (they must protect their membership/income base) and gives them the power to mobilize their members to defend their common interests. They are a self-interested elite, but an elite who's interests are tied to the interests of their following. But if the membership's interests were to change, they would not longer be effective representatives.

Unfortunately, the analysis of leadership in urban social movements has not been subject to sufficient attention to be able to provide the basis of generalization of these conclusions. Such research should, however, be of the utmost importance given the centrality of such movements in the hopes that they will provide the basis for democratization and bottom-up development in the third world.

• 1. Compare, for example, the recommendations of a report in 1990 which would have left most street vendors in the street (DDF 1990) with later proposals for market construction (DDF 1991; 1992: DDF-Secretaria de Gobierno 1992). For a political commentary on the change of attitude after 1991 towards the "popular" sector as a whole, see Monge (1993).
• 2. For example, the number of "Tianguis", periodic markets that often operate in a circuit operating in a different neighborhood each day of the week, increased from 669 to 1061 between 1985 and 1991, while the number of stalls in the markets increased from 84,586 to 152,003 over the same period, despite constant efforts and programs on the part of Coabasto to prevent any growth at all (DDF-Coabasto 1985 and internal reports).
• 3. All the major leaders boast pictures taken at rallies, meetings, and conferences with Federal Deputies, Senators, and in some cases even the President. Partly this is to impress their members, but a large part is also real. Guillermina Rico, the most important leader in the Historical Center, has been actively wooed by the PRI with invitations to "Los Pinos", the Presidential residence, as well as with the position of "Alternate" to the Asamblea de Representantes del Distrito Federal. Several other leaders have also held public office, some for the opposition such as Celia Torres, whose Confederaci¢n de Comerciantes y Organizaciones Populares (CCOP), with an estimated 5,000 tianguis members, was a powerful supporter of the 1988 campaign of Cuauhtemoc C rdenas.
• 4. The political importance of street vendor organizations has been mentioned but not systematically studied elsewhere, for example in Colombia (Bromley 1978), Peru (De Soto 1989), and Hong Kong (McGee 1973). Sanyal (1992) and Escobar Latapi (1990) both provide a general discussion of the political significance of the informal economy as a whole.
• 5. "Fixed stalls" are metal kiosks that typically sell prepared food. "Semi-fixed" stalls are metal frame stalls that can be taken down overnight but are set up each day in the same location. "Toreros" are truly ambulatory vendors, most of whom sell to cars at traffic stops. "Tianguis" are periodic markets that usually operate in a circuit, setting up in a different neighborhood each day of the week. The approximate number of street vendors is a conservative estimate based on data from Coabasto and Mexico-INEGI (1990). See Cross (1993) for an elaboration.
• 6. The argument used by Uruchurtu, and still repeated by officials, is that since street vendors often purchase from semi-wholesalers and even retailers, they must be less efficient commercially since they create a larger number of "intermediates" between the consumer and producers and thus presumably higher prices. However, this overlooked the fact that street vendors are highly competitive with large retail outlets, including supermarkets.
• 7. All information concerning the number and construction dates of markets and concentrations was compiled from DDF-Coabasto (1991). This document lists the date in which markets or concentrations were first "inaugurated" as well as information on the number of vendors they hold.
• 8. Usually, concentrations are located in the poorest areas of the city. Some operate in the street, although most of them have moved to empty lots that they can build on. As noted, some eventually are "re-designated" as public markets. Whether they do so depends as much on the concentration vendors and leaders themselves as upon city officials. Public markets operate under much stricter rules than concentrations, and are administered by city officials rather than by leaders. In many cases, leaders may not allow concentrations to become markets in order to continue exploiting the vendors within them by renting or selling the stalls at high prices.
• 9. Later, however, such vendors became the nucleus of today's powerful street vendor organizations, which helps explain why many of today's leaders are women, and why all of them seem to have such string characters. They have lived through repression and won.
• 10. See Camp (1976), Davis (1994), Hofstadter (1974), Gomez Castro (1963) and Ward (1990) for more information on Uruchurtu's career. For information on the Mexican political system as a whole, also see Camp (1980; 1993), De la Pe¤a (1992), Gonzalez Casanova (1965), Smith (1979), and Story (1986).
• 11. Pyle (1968) pointed to low vacancy rates in many markets to argue that they were a failure.
• 12. Because of its location in public space, a street market is not just a place of private economic transactions: it is a place of socialization, a place to eat, to see and be seen, to chat with neighbors and friends, and a place to see what is available, and what takes one's fancy. On the other hand, public markets are more constraining in the sense that the closed space gives a sense of alien proprietorship, and one must therefore have a specific "purpose" to be there.
• 13. This supposedly is illegal, since the market stalls, like ejido land, were designed to be "inalienable" for the sole benefit of the vendor originally assigned to it. In practice if a vendor found someone who wanted their stall, they could charge them for allowing the stall to be transferred, giving a substantial bribe to the market administrators to change the names on the official roster. In this way, some vendors have amassed a large number of stalls that they purchased from others. Since individuals are only allowed to have one stall, vendors use "name-lenders", usually family members who are listed as stall-holders, but who actually have no control over the stalls at all.
• 14. See Couffignal (1987) and Tomas (1990) for further details on street vending in Tepito.
• 15. A media blitz against vendors in the Metro system and in the Historical Center began at the same time, using negative coverage in newspapers and even prime-time television news shows, which a high official boasted about helping to organize.
• 16. Small stores almost never give official receipts that require the payment of sales and income taxes, and if requested to do so often add a surcharge to cover sales tax even though this is by law included in the original price. Thus, the charge that street vendors are "disloyal competition" because they avoid such taxes could easily be expanded to most of these small "established" stores.
• 17. Later, however, officials noted more optimistically that, while the vendors had refused to purchase the market stalls, the city had been able to push them off their original positions on the surrounding streets. Several officials later raised the same claim with regards to the entire Historical Center project.
• 18. Officially, the market stall itself, as a form of condominium real estate, was the only form of security for the individual loans that vendors incurred. For a few markets that were located on Metro property that could not be alienated, vendors were not owners but merely lessees, and the lease itself had to suffice as the security.
• 19. Newspaper and book stalls have special protection under Mexican law and are jealously protected by publishers. They are also regulated under a different code for service categories, which includes shoe shiners and street musicians as well.
• 20. One young man caring for a child at La Merc‚d said he never sold more than three pairs of sandals a day, but his wife, the real owner of the stall, sold illegally in the nearby streets to pay the mortgage and their living expenses. An elderly man who spoke eloquently about his attempts to convince his fellow vendors to give their market (in another area) a chance to build up a clientele actually earned most of his money a block away from the market on a pedestrian bridge. An Indian woman with a fruit stall in a small but empty market near La Merc‚d noted that she had to sell her fruits from door to door and on the street to keep her merchandise from rotting in the market.
• 21. In San Ciprian vendors noted that several groups had begun to buy up stalls in the market, but were keeping the stalls closed. The most probable reason is that by so doing, they kept the value of the market low, since more open stalls would attract more customers. Keeping the stalls closed decreased the commercial value of the market as a whole and thus the resale value of the stalls the buyers were after, as well as pushing the small vendors into bankruptcy. It must be remembered that, while many markets were failures for small vendors because they were often poorly designed and advertised, they are all located in prime commercial and business areas in the center of the city where land values are potentially astronomical.
• 22. Vendors at one market claimed that they had been required to sign over 50 documents without reading them before being assigned stalls.
• 23. Indeed, a nationwide movement began at this time against the banking sector accusing it of usurious interest rates. Known as the "Barz¢n", it included people from all ranges of life, but its existence also reflected the results of Salinas' neo-liberal policies that had persuaded more Mexicans than ever to go into debt. In is certain that the vast majority of the street vendors who signed mortgages for stalls in the Historical Center had never held formal loan papers before.
• 24. It should be pointed out that the most successful markets, Pino Suarez and Meaves, were both planned long before this period at the instigation of the vendor organizations themselves, which were far more active in the design and construction process than other groups.
• 25. Leaders became a sort of class of political entrepreneurs, since their power and income was based on the number and quality of vending areas they controlled. Since space was controlled by the organizations, they also controlled the permit system--nobody could have a permit without being a member of an organization.

• Baca, Pedro
  1990 "Las Zarinas de las Banquetas." in Contenido. Mexico City. August, 1990.
• Bromley, Ray
  1978 "Organization, Regulation and Exploitation in the So-Called `Urban Informal Sector': The Street Traders of Cali, Colombia" in World Development 6:1161-1172.
• CANACO
  1987 La Econom¡a Informal. Mexico, D.F.
• 1990 Econom¡a Informal, Tercera Edici¢n. Quien Provee a los Ambulantes. Mexico, D.F.
• Camp, Roderic A.
  1976 Mexican Political Biographies: 1935-1975. The University of Arizona Press. Tuscon.
• CETEPI (Centro de Estudios Tepite¤os)
  1991 El Ambulantaje en el Centro Historico. Unpublished report.
• Cornelius, Wayne A.
  1975 Politics and the Migrant Poor in Mexico City. Stanford University Press, Stanford.
• Couffignal, Georges
  1987 "Misterioso Tepito". Trace, No. 11.
• Cross, John C.
  1992 "Studying Streetvending in Mexico City" in Mexico Policy News, #8, Fall 1992. pp 17-18.
• 1993 The Politics of Street Vending in Mexico City: The Other Resistance. University Microfilm International. Ann Arbor, Michigan.
• DDF
  1990 Diagnostico y Propuesta de Ordenamiento del Comercio en la Via Publica del Centro Hist¢rico de la Ciudad de Mexico. Mexico, D.F. March, 1990.
• 1991 "Comercio en Via Publica, su Reubicacion, una Decision Politica". Unpublished internal DDF Document. August, 1991.
• 1992 "Programa Inmediata de Mejoramiento del Comercio Popular". Unpublished internal DDF document. August, 1992.
• DDF-Coabasto
  1985 "Proyecci¢n del Numero de Tianguis para 1988". Unpublished internal document. April 25, 1985.
• 1991 Directorio de Mercados P£blicos y Concentraciones. Unpublished internal document.
• DDF-Cuauhtemoc
  1990 "Informe de Actividades del C. Lic. Ignacio V zquez Torres, Delegado en Cuauhtemoc" (presented to ARDF June, 1990.)
• DDF-Secretar¡a General de Gobierno
  1992 "Programa de Mejoramiento del Comercio en la Via Publica". Unpublished internal Document. February, 1992.
• De la Pe¤a, Guillermo
  1992 "Una Nueva Cultura Politica?" in Alonso, Jorge, Alberto Aziz and Jaime Tamayo (eds) El Nuevo Estado Mexicano. IV. Estado y Sociedad. Mexico. Nueva Imagen.
• De Soto, Hernando
  1989 The Other Path: The Invisible Revolution in the Third World. Harper & Row, New York. Translation from Spanish (Original 1986)
• Eckstein, Susan
  1977 The Poverty of Revolution: The State and the Urban Poor in Mexico. Princeton, N.J.: Princeton University Press.
• Escobar Latapi, Agust¡n
  1990 "Estado, orden politico e informalidad: Notas para discusion" in Nueva Antropologia 40:23-40.
• Excelsior. Daily Newspaper. M‚xico, D.F.
• Fernandez-Kelly, M. Patricia & Anna M. Garcia
  1989 "Informalization at the Core: Hispanic Women, Homework, and the Advanced Capitalist State" in Alejandro Portes, Manuel Castells & Lauren A. Benton, eds. The Informal Economy: Studies in Advanced and Less developed Countries. The Johns Hopkins University Press. Baltimore, Maryland.
• G¢mez Castro, Arturo
  1963 Qui‚n Ser el Futuro Presidente de M‚xico? Editorial Veracruz, M‚xico, D.F.
• Gonz lez Casanova, Pablo
  1965 Democracia en M‚xico Ediciones ERA, M‚xico, D.F.
• Hofstadter, Dan, Ed.
  1974 Mexico 1946-73. Facts on File, Inc. New York.
• La Jornada. Daily Newspaper. Mexico, D.F.
• La Prensa. Daily Newspaper. Mexico, D.F.
• Lerner de Sheinbaum, Bertha & Susana Ralsky de Cimet
  1976 El Poder de los Presidentes: Alcances y Perspectivas (1910-1973). Instituto Mexicano de Estudios Pol¡ticos, M‚xico, D.F.
• Marx, Karl
  1963 The 18th Brumaire of Louis Bonaparte International Publishers, New York.
• McGee, T. G.
  1973 Hawkers in Hong Kong. Center for Asian Studies Monographs and Occasional Papers, No. 17. University of Hong Kong.
• Mexico
  1951 "Reglamento de Mercados" (Del Distrito Federal) in Diario Official. June 1, 1951. pp 6-15.
• Mexico-INEGI
  1990 Encuesta Nacional de Econom¡a Informal. Mexico. INEGI.
• Monge, Raul
  1993 "No tienen cabida los pobres en la utopia urbana de Camacho" in Proceso #864, May 24, 1993. pp 10-13.
• Parnell, Philip C.
  1992 "Time and Irony in Manila Squatter Settlements" in Nordstrom, Carolyn & JoAnn Martin (eds) The Paths to Domination, Resistance, and Terror. Berkeley. University of California Press.
• Portes, Alejandro & John Walton
  1981 Labor, Class and the International System. New York: Academic Press.
• Pyle, Jane
  1968 The Public Markets of Mexico City. University Microform Library Service. Ann Arbor.
• 1978 "Tianguis: Periodic Markets of Mexico City" in Robert H. T. Smith, ed. Marketplace Trade--Periodic Markets, Hawkers, and Traders in Africa, Asia and Latin America. Centre for Transportation Studies. Vancouver.
• Ramirez, Miguel D.
  1986 "Mexico's Development Experience, 1950-1984" in Journal of Interamerican Studies and World Affairs 28:39-66.
• Sanyal, Bishwarriya
  1991 "Organizing the Self-employed: The Politics of the Urban Informal Sector" International Labor Review 130:39-56.
• Smith, Peter
  1979 Labyrinths of Power: Political Recruitment in Twentieth-Century Mexico. Princeton. Princeton University Press.
• Story, Dale
  1986 The Mexican Ruling Party: Stability and Authority. Praeger. New York.
• Tokman, Victor E.
  1978b "Competition between the Informal and Formal Sectors in Retailing: The Case of Santiago" in World Development, 6:1187-1198.
• Tomas, Francois
  1990 "Tepite¤os". Trace, No. 17.
• V zquez Torres, Ignacio
  1991 El Abasto en la Ciudad de Mexico. Mexico, D.F. COABASTO.
• Wall Street Journal. Daily Newspaper. New York.
• Ward, Peter
  1990 Mexico City: The Production and Reproduction of an Urban Environment. London. Bellhaven Press.