Private Education Supply in Disadvantaged Areas of the City of Buenos Aires and ‘ Low-fee Private Schooling ’ : Comparisons , Contexts , and Implications

Recent studies have revealed an increase in the enrollment of students coming from vulnerable social sectors in private primary schools in the City of Buenos Aires. To date, the phenomenon of ‘private education for the poor’ has been principally studied in Asian and African countries, where the deficit of state educational supply has given rise to a multiplicity of private education undertakings of diverse nature. Using official statistical data, this article explores and describes the structure of private education supply in area s of low socioeconomic status in Buenos Aires and assesses the extent to which the category of low-fee private school (LFPS) contributes to understanding this phenomenon. Our analysis reveals a great heterogeneity in this sector of private education in Buenos Aires with regard to financing structures, monthly fees, and the legal forms in which schools operate. We also find the presence of a significant religious component. We suggest the possibility that private provision has helped to mitigate coverage imbalances at a 'microlocal' level in some of the studied areas. Our comparative approach evidences that the notion of LFPS provides a more normative than descriptive view, limiting the perception Education Policy Analysis Archives Vol. 23 No. 126 2 of a phenomenon that is actually much broader, multifaceted and dynamic.


Introduction 1
Recently, private elementary schools have become an option for many low-income families in the City of Buenos Aires who traditionally used to send their children to state-run schools (Gamallo, 2011;Krüger, 2014).From the mid-1990s to the early 2000s the presence of such families in the private sector was marginal and only middle and high-income families opted out of public schools (Gómez Schettini, 2007).Nevertheless this reality has undergone change.In 2012, for instance, nearly a third (28.7%) of children in the poorest income quintile were attending religious and non-religious private elementary schools.This article provides a description of the main features of this expanding sub-sector of private education in Buenos Aires.In addition, we raise the question of whether these schools may be considered part of the so-called 'low-fee private schools' (LFPSs) phenomenon that has been growing in many low and middle-income countries as a response to the absence or perceived inadequacy of public education (Akaguri, 2014;Balarin, 2015;Dixon, 2012;Härmä, 2011Härmä, , 2013;;Phillipson, 2008;Srivastava, 2007Srivastava, , 2013;;Verger, forthcoming, among others).
To gain insight into the LFPS phenomenon a provocative article by Walford (2011) put to the test a comparison between Christian private schools in the UK and LFPSs emerging in Asia and Africa.Interestingly, Walford notes that despite the contextual differences, British Christian schools show many similarities with respect to LFPSs operating elsewhere.This comparative approach is invaluable in a variety of scenarios.It can help to understand the structural interdependence between the public and private sectors, and how working-class families behave facing the public-private alternative in very different contexts.
In this vein, we argue that a comparative benchmarking exercise is relevant to analyze the private schools in the City of Buenos Aires that serve disadvantaged families.The existence and heterogeneity of these schools have remained largely unexplored.Naturally, there are differences between these schools in Buenos Aires and LFPSs elsewhere as we will see.However, there are important structural similarities that need to be considered, namely, the students' sociological status, the private nature of schools, their role in contexts where public education is unavailable or considered inadequate, and implications regarding social justice.
Furthermore, Buenos Aires has the highest private enrollment share in Argentina and one of the highest in Latin America with more than 50% of children attending private schools.The expansion and heterogeneity of the private sector has been driven by 'historic public-private partnerships' in the form of state subsidies to private schools since the 1940s (Verger, Zancajo, & Fontdevila, Forthcoming;Narodowski & Moschetti, 2015).These subsidies have enabled a special type of private school to emerge that establishes substantially lower fees than 'traditional' private schools.Such fees are affordable for low-income families -even when in relative terms they do represent a substantial part of the household income.Interestingly, LFPSs elsewhere are increasingly being integrated in public-private partnership schemes as part of what Srivastava (2015) calls 'the global scaling-up of low-fee private schooling'.These features make the case of the city of Buenos Aires particularly relevant to understand how these initiatives could evolve.
Finally, besides exploring the main features of this sub-sector in Buenos Aires and analyzing its role in a broader educational structure, another aim of this article is to provide insights into the usefulness of the notion of LFPS as a theoretical category.
The article is organized as follows.In section two we provide background to the LFPS model, thus setting the framework to analyze the Buenos Aires case.We focus on the three main areas of discussion: LFPS definition and implications in understanding broader privatization processes, the regulatory environments in which LFPSs operate, and quality and equity related concerns.
In section three we present an overview of the privatization process undertaken by the Argentine education system in its recent history.We then discuss the general characteristics of both the private education sector of the City of Buenos Aires and the subsidy allocation system emerging from a historic public-private partnership.In section four, drawing on official statistical databases, we build and analyze a series of indicators intended to map out first the distribution of enrollments across communes2 and by income quintile, and then to characterize private schools in the city's communes 4, 7, 8 and 9, which gather most of the highly vulnerable population in the city.Finally, a brief concluding section summarizes the central ideas of the article.

What are LFPSs and Where Do They Stand in Privatization Processes
The LFPS model is not confined to a single definition as its great proliferation has led to numerous ventures, adapted to the characteristics of each area.One of the biggest problems in defining LFPS is determining what 'low-fee' means.In this respect, Verger (Forthcoming) emphasizes that this concept is contingent on not only 'the social structure of the countries where these schools operate, but also on the economic circumstances of particular families'.Srivastava (2008, p.453) tries to set a threshold and points out that in order to qualify as low-fee, schools must not charge a monthly fee exceeding the amount corresponding to the daily income of a nonqualified worker, for primary level, and no more than twice that amount for secondary level.However, these precise statements are problematic when considering cases where some LFPSs are partly subsidized by the state, as occurs in Nigeria or South Africa (Härmä, 2013;Schirmer et al, 2010;Ahmed & Sayed, 2009).
In a study on LFPSs in Asia and Africa, Phillipson (2008) defined them as 'schools created by an individual or group of individuals aiming at financial profit' (p.1), and excludes any school operated by nonprofit organizations, religious organizations, community organizations, and of course, traditional private schools.This definition has some problems that have been identified by Walford (2011).In fact, Phillipson's definition is an extreme simplification of the reasons why private schools are created to serve the low-income population.Walford argues that with the exclusion of low-fee schools run by NGOs (Non-Governmental Organizations) and religious organizations, we lose sight of much of this new trend of privatization that differs from traditional privatization modes, and he contends that the very idea of profit is relative: [the definition] restricts our understanding of why such schools might be started and how the schools themselves, and the motivations for their continued existence, may change in nature over a period of time.For example, even the idea that they must 'make a profit' collapses within the complexity of individuals and groups paying themselves salaries, or establishing schools so that they might gain employment.What is clear is that the vast majority of these new private schools are not the result of shareholders investing money in schools because they see that as the way to obtain the highest financial return.The reasons for starting and continuing with the schools are much more complex, and there is thus the need to consider the whole range of non-government-sector schools with low fees that are designed to serve some of the poorest families in each society.(2011, p. 403) Underlying the problem of defining LFPS is a debate about what type of privatization thei r appearance and development imply.Returning to the limited definition offered by Phillipson (2008), we find an explanation for LFPSs that hails market mechanisms as a solution to the problems of state supply deficit.Along these lines, Tooley (2013, pp. 447-448) considers it is a process of 'grassroots privatization', thus adding to the equation a component of supposed spontaneous community organization.He views this as inherent to the LFPS model, following a 'bottom-up' type of logic regardless of the state.However, as Verger (Forthcoming) notes, these postulates become contradictory when they are immediately followed by recommendations in favor of governments and international organizations to help with funds and encourage this type of school provision.Do these schools then lose their LFPS status as soon as they begin to receive public funding and somehow become part of state educational policy?
In short, the LFPS model is a way of privatization that, Verger notes (Forthcoming), began as a phenomenon resulting from the state's failure regarding school coverage and/or quality; namely a kind of default privatization.But today it takes more complex forms, due to the involvement of international organizations and states themselves in the shaping of public-private partnerships (PPPs).Indeed the LFPS model is increasingly being proposed as a valid education policy alternative.Its financial model may follow traditional patterns of partial public funding to private schools, considered as an emerging level PPP (Verger, 2012).Alternatively, as put forth by some researchers (Dixon, 2012;Tooley & Dixon, 2003), it could involve directing funds from international organizations to finance these schools, or establishing education voucher systems to increase opportunities for school choice.In this respect, the debate over the definition of the LFPS model is far from conclusive.As noted by Junemann and Ball (2015), determining what is LFPS demands studies that reveal the specifics of each private form of school provision aimed at the most vulnerable social sectors.

Problems of Acknowledgement and Legal Status
Most researchers consider LFPS as a phenomenon that arose in the shadow of deficient state education systems.In the documentation about LFPSs, there are schools created for profit by their founders, and many others resulting from more altruistic motives (Walford, 2011).Nevertheless, they share their origin as a response to the absence or to the perceived inadequacy of the state provision.Accordingly, Härmä (2009;2013) highlights that those families who choose these schools never base their decision on an ideological overestimation of the private mode.On the contrary, the decision and choice process in favor of private supply is eminently pragmatic and is analogous to what happens when the state fails to adequately provide other public supplies and services, either in quantity or in perceived quality.
As regards the increasing number of LFPSs, the reaction of states seems to drive schematically to two poles.On the one hand, in most cases, the trend is to grant no official acknowledgement to these schools and as a result, their activities become simply illegal.However, in other cases, some of these schools actually get official acknowledgement either because they meet the corresponding legal requirements, or by virtue of what Srivastava (2008, p.468) calls 'shadow institutional framework', in the case of India.In this case, the authorities grant acknowledgement to LFPSs, even though they do not meet the current regulatory framework.In Srivastava's point of view, this acknowledgement occurs as a result of pressure from two sources: first, awareness on the part of the authorities regarding the official regulatory framework containing a number of perverse incentives that force private schools to meet requirements that public schools never achieve.Second, awareness that LFPSs unquestionably contribute to increase school enrollment rates where the state cannot expand the coverage.This is where the juxtaposition of the goals of Education For All and the budgetary restrictions of developing countries lead states themselves and international organizations like the World Bank to begin perceiving LFPSs as valid ways to extend coverage (see Verger and Bonal [2012]).

Quality and Equity
Even when LFPSs actually increase coverage in countries where the state fails to expand free schooling, questions inevitably emerge regarding quality and equity.The evidence concerning the quality of these schools is not conclusive at all (Junemann & Ball, 2015;Day Ashley et al., 2014).Several studies claim that the performance of students, the infrastructure and the teaching staff of LFPSs are superior if compared to the available state supply (Dixon, 2012;Garg, 2011;Tooley, 2013;Tooley & Dixon, 2003, 2006).However, these findings are relative and limited: state supply shows a heterogeneity that is not considered in these studies and that is by no means negligible.Moreover, the quality indicators used are not rigorous enough, the scale of these results does not support generalization, and relevant contextual elements are ignored in the analysis.Sarangapani and Winch (2010) present one of the strongest objections: Yet in the current study, Tooley et al. have chosen a new set of very unusual metrics as proxies for 'quality' of educational opportunities.The authors are correct in claiming that basic health and hygiene, safety and comfort and the presence of teacher activity are all important.But the extensive data gathered, tabulated and analysed, of various infrastructural aspects, including the presence of fans, tape recorders and televisions, on which they conclude that the PUU are better or at least as good as government schools, borders on the absurd, particularly when taken with other possible indicators which are omitted.Indeed, they choose these relatively trivial indicators over other proxies such as teacher training qualifications or time-tabled activities!(2010, p. 511) Some researchers also question the 'inclusive' nature of LFPSs, observing that no matter what the cost is, it is still unaffordable for the poorest families (Akaguri, 2014;De, Noronha, & Samson, 2002;Fennell & Malik, 2012;Härmä, 2011;2009).This reality raises a debate about the equitable nature of LFPS development.Additionally, the supposed cost-efficient nature of lowcost undertakings often goes hand in hand with low quality in teaching, perhaps the most important component of a quality educational experience (Härmä, 2011, p. 563).Regarding equity issues and these schools possibly substituting a free but inadequate state supply, Tooley (2013) makes a theoretical effort to redefine the concept of transcendental Rawlsian justice by resorting to Amartya Sen's notion of 'relative' or 'possible' justice.He arrives at the conclusion that LFPSs are not necessarily unfair, or at least not more unfair than the status quo.Tooley's stance is interesting because it tries to dismiss the value of a supposed 'egalitarian utopia' and paves the way towards a typically market pragmatism.However, even if we exclude from the analysis those families that cannot afford even a low-cost school -that in Tooley's scheme, could benefit from some voucher-like or affirmative action program-and we consider only equity issues concerning low-income families that can actually afford to pay for these schools, the questions raised by Härmä persist: Diverting poor families' spending in this way from other areas of essential expenditure may prevent or delay their movement out of poverty, and may prevent much needed investments in small businesses and other livelihood opportunities.Allowing this situation to continue is manifestly against the interest of working towards a more equal and just society.(2013, p. 550) In this section we have briefly discussed the main areas of academic and policy concern regarding the LFPS phenomenon.In order to undertake a comparative benchmarking exercise we have focused on three main topics that provide a framework for analyzing private schools that serve disadvantaged families in Buenos Aires.We have also raised some questions and noted some issues still under debate concerning the nature and workings of LFPS that need to be kept in mind for any analysis of this sector.We will draw on this in the discu ssion section of the paper.
The Privatization of Education in Argentina: New (and not so New) Trends

Partial Privatization of Education Supply
During the second half of the twentieth century, the Argentine education system underwent a transformation defined by sustained growth of student enrollment in private schools.This process, which started in the 1960s, broke with the existing arrangement of education provision.Prior to that moment, education in Argentina was eminently state-run, via a process of 'statalization' of the preexisting supply as occurred in many countries between the XIX and XX centuries (Narodowski, 2011;Perazza, 2011).Demand-side pressures, deregulation policies and changes in the financing structure of the system initially powered the transformation in the 1960s within a logic framework that some consider as quasi-spontaneous (Morduchowicz et al., 1999) or by default (Narodowski, 2008).The present Argentine education system reflects a 'quasi-state monopoly' structure resulting from a dynamic balance between the traditional state sector and a private sector of great heterogeneity, partially financed by public funds (Narodowski, 2011).Within the context of chronic fiscal shortages, habitually emerging in developing countries, the private sector is a cost-efficient alternative that can enable maximization of school coverage.Families that can finance their exit to the private sector 'release' state funds and vacancies in public schools.In turn, the associated cost of a scheme with these characteristics has been the increasing, well-documented socioeconomic homogenization within each subsector (Gasparini, Jaume, Serio, & Vazquez, 2011;Narodowski, Moschetti, & Gottau, 2013).
In terms of the dynamics between public and private provision, enrollment in private schools seemed to held steady at approximately 24%3 in the mid 1990s (Narodowski & Moschetti, 2015).However, despite the profound economic crisis that the country underwent during 2001-2002 4 , and the post-crisis government's discourse in favor of public education, since 2003 the private education sector's growth resumed and reached 28.9% in 2013 (DiNIECE, 2014) 5 .This transformation may be related to the economic recovery of the middle classes throughout the period.However, this sole explanation cannot account for the phenomenon in all its magnitude.The Argentine education system is federally organized and each of the 23 provinces as well as the Autonomous City of Buenos Aires is responsible for the provision, regulation and financing of their systems.The group including the provinces of Buenos Aires, Córdoba, Santa Fe, and the Autonomous City of Buenos Aires account for 59% of total enrollment in compulsory education.In the Latin American educational context, Argentina ranks second in privatization, only outranked by Chile whose education system is, unlike the Argentine, structured on the basis of demand-side financing and a quasi-market model (Narodowski & Nores, 2002;UNESCO, 2014).If we consider variations by province for the period 2002-2013, the growth trend is consistent throughout all Argentine provinces, though major urban centers and the most populous provinces show more private participation (Moschetti, 2013). .
In the City of Buenos Aires, while 47.8% of students attended a private school in 2002, 50.7% did in 2013, and so private school has become the main mode of provision in the district for levels included in compulsory education (DiNIECE, 2014).Studying the case of the City of Buenos Aires is interesting due to the unprecedently high participation of the private sector in the provision of education.Moreover, recent changes in the characteristics of the population served by these private schools add to the relevance of this case.According to data from the 2012 Annual Household Survey of the City of Buenos Aires for primary education, 28.7% of children in the poorest quintile attended religious and non-religious private schools.For the second poorest quintile, the percentage rises to 46%.Additionally, the survey reveals that 13.7% of primary school children living in villas miseria 6 or very poor households attend privately-run schools (Narodowski, Rozada, Moschetti, & Gottau, 2015).These data are very striking in a context like Argentina where the public school system had in the 1950s around 93% of total enrollments.Argentina had at the time the highest rates of school attendance in the region: by 1955, 79.9% of the population aged 14 to 17 had completed primary school and the primary school net enrollment rate in 1960 was 85.6% (Morduchowicz et al., 1999;Rivas, Vera, & Bezem, 2010;SITEAL, 2014).

Partial Public Funding
The process of private school enrollment growth was accompanied in all provinces by an increase in state funding aimed at partially financing teachers' salaries in most private schools 7 (Moschetti, 2013).These subsidies respond to a dispositif of state support to the private sector that has existed in Argentina since the 1940s.Its logic respects in practice the idea of 'public funding and private supply', considered the guiding principle of every public-private partnership (Verger & Bonal, 2012).Given its longstanding existence, some have called this kind of scheme 'historic public-private partnerships' (Verger, Zancajo & Fontdevila, forthcoming). 6Villas miseria are urban agglomerations of informal, densely populated settlements composed of extremely substandard housing.In the City of Buenos Aires, there are 20 villas miseria, mainly gathered in communes No. 4, 7, 8, and 9. See Kessler andDi Virgilio (2008), andClichevsky (2003). 7According to a study by Mezzadra and Rivas (2010), 65% of private institutions of the country benefit from Sstate funding to some extent.In the case of the Province of Buenos Aires, there is some progressivity between the proportion of granted subsidies and the income level of the served populations.However, the study also evidences significant inconsistencies: '17% of schools within the poorest tercile receive less than 50% of State subsidy.In turn, 16% of schools within the highest tercile receive more than 75% of State contributions' (Mezzadra and Rivas, 2010, p. 58).Except for this study regarding the Province of Buenos Aires, there are no data on how subsidies are distributed in each province.
Since the 1960s, in addition to state funding, the expansion of the private sector was driven by a process of deregulation with respect to the bureaucratic hierarchies of the state sector.Along this process, spontaneous changes in preferences by those demanding education were made easier or induced by new regulation arrangements (Morduchowicz, 2005, p. 39).These new regulations and deregulations formed a legal framework that would guarantee the activities of private institutions, thus ensuring the stability and acknowledgement of teachers in the sector, the validity of certificates issued by non-state institutions, in addition to regular allocation of public funds in order to afford teacher salary expenses in some proportion (Morduchowicz, 1999).There is no evidence to exclusively attribute the growth of the sector to the systematic granting of funds.Nevertheless this practice has provided incentive and validation.
The system of state funding to private education originated in 1947 under the first presidency of Juan D. Perón (Law 13,047).The purpose was twofold: to ensure school provision in areas where there were no public schools, and to supply financial aid to private schools-mostly catholic schools-serving low-income households.That event marked the beginning of the systematization of a policy that subsequently, and through successive decrees (Nº 15 of 1964, and Nº 2,542 of 1991), gained a clearer definition regarding the proportion of teacher salary expense that the state would finance (currently from 40 to 100%).The basis used to calculate the amount for each school has always been teacher salary expense, considering only those who teach contents included in the official curriculum.The norm also establishes certain criteria to define grants and their proportion (socioeconomic status of the pupils' families, proximity and availability of similar state schooling, etc.).But these are not comprehensive, and the decision rests, as some studies note, on good judgment and at the discretion of the relevant authorities (Mezzadra and Rivas, 2010;Morduchowicz 1999).Public funds thereby do not compensate for real estate investment, extracurricular teacher salaries or other related expenses.

Distribution of Enrollment and General Indicators
The education system of the City of Buenos Aires is heterogeneous, including a private sector of great quantitative relevance and of great variety.Historically, it has been private schools that have provided religious education, mainly Catholic, but also Protestant, Jewish and Islamic.In addition, there are several schools that reflect the immigration history of the country (Italian, Spanish, British and Armenian schools, among others).Further, there are others that stand out by offering certain pedagogical orientations that are not feasible within the frame of hyper regulation the state provision is subject to (for instance, Waldorf, Piagetian, Art Education, etc.) (Narodowski, 1998).The picture gets even more complex if we take into account that many of these schools receive some kind of state funding.
Traditionally, as in other Latin American metropolitan areas, private schools in Buenos Aires were considered to serve exclusively middle and high-income households and it was the middle class that benefited the most from subsidized lower monthly fees (Pereyra, 2008).However, a detailed current view into the system adds new elements of analysis and introduces some nuances into this scenario.Table 1 presents the distribution of primary school enrollment among public schools, private schools, and private religious schools8 in 2012, for each of the 15 communes of the city.Even in those communes that concentrate most of the low-SES households, the students' enrollment rate in private schools ranks from 25 to 43%.Most of the city's population living in villas miseria and other precarious settlements is located in these communes (4, 7, 8 and 9).In these places, family income per capita is lower than the city's average by more than 10%, and socio-educational indicators are the most critical of the district (DGECE, 2014).If in addition we take a look at the distribution of the enrollment by income quintile, the participation of the private sector within the two lowest quintiles is considerable (Table 2).Moschetti, & Gottau (2015).
Even though we note that throughout the whole City religious private education has a stronger presence than non-religious private education, in low-SES communes the gap in favor of religious education widens.A priori, we could consider that this pattern is related to the fact that religious schools are often the main recipients of state subsidies and therefore, the ones charging the lowest fees.However, such a generalization is not entirely correct: It should be noted that, on the one hand, in commune 8 (low-SES), private enrollment is divided almost equally between religious and non-religious schools; on the other hand, in communes 13 and 14 (high-SES), private religious schools account for a very high share.These elements indicate the ubiquitous and heterogeneous nature of private religious education in the City.There are religious schools at all levels of the 'local hierarchies'.Some offer education at prices only accessible for the most well-off, while others charge subsidized (or highly subsidized) monthly fees that benefit the middle and low-income sectors.

Supply Structure in Communes 4, 7, 8 and 9
We now characterize how supply is structured in these contexts in terms of public and private provision, identifying how many private institutions benefit from state financial support and to what extent.It is also of interest to know whether education run by religious and non-profit organizations is as prevalent as is thought.What is ultimately the role of private schooling in these areas?Does it contribute to mitigate a problem of public education shortage?
If we look at the structure of supply by commune, in particular as regards the situation in communes 4, 7, 8 and 9 (Table 3), we notice a number of issues that deserve some further consideration.First, low-SES communes show the highest number of students and sections9 in the City.In the case of communes 4 and 8, the number of enrolled students is virtually twice the number of students of in other communes.Second, public schools have a greater number of sections per school than private schools.Third, on average, the number of students per section in both public and private schools is slightly higher in 'poor' communes.Finally, the number of students per section in private schools is consistently higher than the number of students per section in public schools in all communes.
Even though the poorest areas of the city gather a larger number of students due to their own socio-demographic dynamics10 , the data presented show no evidence, in principle, of overcrowding in public school classrooms.From the aggregate perspective that these indicators make possible, the system appears to be balanced, and the expansion of private schooling has certainly played a role in achieving full school attendance rates.Nevertheless, overcrowded school scenarios have in fact been observed in many public schools in low-SES communes (Martinez, 2012).Here, the commune as a territorial unit includes urban heterogeneities that tend to average out and mask the dynamics that occur mainly at a 'micro-local' level.These dynamics are of greater complexity in communes 4, 7, 8 and 9 where there has been overwhelming growth of villas miseria and informal settlements.This expansion followed a very different timing and growth pattern with respect to predictions regarding urban planning.Between 2001 and 2012, the population living in these settlements in the City increased by 81%, from 107,422 to 194,228 (Mazzeo, 2013).Consequently, the very location of schools is a problem and the system will not always be able to respond 'in balance' to a demand that reveals itself as constantly changing.Balance in general terms does not imply in any way the absence of imbalances that may be found at a more local scale than the commune (i.e., neighborhoods).Full school attendance rates reveal, therefore, that private provision has contributed to the overall balance, but also suggest that it has helped to mitigate coverage imbalances unevenly across neighborhoods that would deserve in-depth study.
A second point to note is that in all cases private schools have substantially more students per classroom than public schools and that this trend occurs more noticeably in poor communes.At the same time, public schools tend to have more sections per school, either because school buildings have on average higher capacity, or because the half-day school mode -of greater relative importance in public schools than in private schools (see Tuñón and Halperin [2010])-enables the possibility of opening a section in the morning and another one in the afternoon, which would count for one section at a full-day school.
Related to the above, our data is the consequence of two different strategies of 'efficient' use of resources, both with significant implications in terms of quality.First, maximizing the number of students per section in an attempt to maximize the use of classrooms; and second, by means of halfday school modalities, increase the number of sections so as to maximize the use of the school building.Commune 8 is the most extreme example of these dynamics, with 22 sections per public school and 31.7 students per section in private schools.These initiatives, which acquire particular characteristics depending on the possibilities of each context, make full coverage in primary education feasible, but at the same time cause strong imbalances and inequalities between schools.With 31.7 students per section, private schools in commune 8 can hardly provide quality education in the same fashion as private schools in middle or high-SES communes where the average is 23 students per section11 .
Looking at the structure of the private sector in communes 4, 7, 8 and 9, the heterogeneity of legal forms adopted by schools is striking (Figure 1).Here again we find a complex pattern of interdependencies between public and private -in its various forms-that according to Walford (2011) makes a restrictive definition of LFPS an oversimplification of limited use.Moreover, there are no legal restrictions in most Argentine provinces for educational institutions to be for profit.In the City of Buenos Aires, both for-profit and non-profit schools are entitled to apply for state subsidies ranging from 40 to 80% of teacher salary expense.This certainly contributes to the heterogeneity of legal forms we find across the city's communes.Only those institutions that charge no monthly fees are entitled to apply for a subsidy representing 100% of teacher salary expense, and only in these cases profit is not allowed (Decree Nº 2,542/91).Regarding the financing structure of these institutions, only 10% do not receive state funding.The remaining 90% receive state funding ranging from 40% to 100% of teacher salary expense.In most cases -78%-schools receive a contribution greater or equal to 80% (Table 4).The systematic use of these supply-side funding mechanisms in order to meet educational needs outside the domain of the state sector is clearly demonstrated here.Considering the supply structure in these communes according to the type of religion of schools (Table 5), we confirm part of what we expected: Religious schools generally benefit the most from public financing.This is mainly due to historical reasons.The country's 'historic public-private partnership' scheme was in part based on the need to integrate the preexisting catholic schools into the national education system.Unlike what happens in more traditionally secular states, in Argentina the fact that religious institutions receive state subsidies is a longstanding and somewhat naturalized phenomenon stemming from the legitimacy of the 'teaching freedom´ principle (Verger, Zancajo & Fontdevila, Forthcoming;Morduchowicz, 2005).In the case of communes 4, 7, 8 and 9, there are no religious schools that do not receive subsidies (unlike what happens in other communes).However, it is important to highlight here that the legal form of schools does not automatically determine whether they receive subsidies or not.There are schools registered as 'Public Limited Companies' and 'Limited Liability Companies' that benefit from subsidies, among many others of the same legal form that do not.Similarly, there are 'Non-profit Organizations' in both groups.
As we saw, one of the controversial points about defining the LFPS category lies in the difficulty of establishing a fee threshold.In this respect, we emphasized the importance of contextualization.In the case of low-SES communes, the average monthly fee of non-subsidized schools is AR$ 2,203 and AR$ 71712 in subsidized schools.Figure 2 provides relevant data showing the number of schools (% of total) for each fee interval.There are 41 schools that charge a monthly fee of less than USD 40 13 , including some with fees even lower than USD 20.These fees represent between 5% and 12% of the minimum wage in AR$ 4,400 (National Ministry of Economy, 2014), although it is evident that the comparison is valid only for formal workers.

Discussion
This study shows the heterogeneous structure of the private education sector within the most vulnerable areas of Buenos Aires.This diversity is evidenced by different legal configurations, religions, bonds with the state system and monthly fees that each school charges, among other variables.The heterogeneity observed has also been reported in countries where the study of LFPSs originated (Härmä, 2013;Srivastava, 2008;Heyneman & Stern, 2013).In line with Junemann and Ball (2015), we argue that such variation in practices calls for reconsidering the categories to analyze private education in disadvantaged contexts, as well as the delimitation and implications of the phenomenon.
In the case of Buenos Aires, schools have a substantial religious component, which is greater in disadvantaged communes than in those of middle or high socioeconomic status.This religious component has a certain independence from the legal forms under which these schools operate.There are religious schools operating as part of religious congregations, as NGOs, foundations and limited liability companies, among others.In turn, there are for-profit institutions, though the vast majority fall into schemes that only aim at covering operating costs: 54.3% of schools in communes 4, 7, 8 and 9 receive subsidies representing 100% of teacher salary expense.According to current legislation, in order to obtain such an amount of subsidy, schools must not pursue profit.Among the schools that receive less than 100% subsidy on teacher salary expense, 32% are NGOs or foundations, which are by definition non-profit organizations.
We have seen that the for-profit/nonprofit dichotomy is not always sufficient to describe the ways in which institutions can operate from an economic and financial perspective.However, the purest definition of LFPS requires that, in order to qualify as such, a school must pursue profit (Phillipson, 2008).We argue that this definition is more normative than descriptive.The creation and use of the concept of LFPS in policy rhetoric and discourse seems to stem from the determination to put forth a stable, standardized and replicable model of education, alternative to the state model, based on only some of the many forms of educational provision that spontaneously arise bottom-up outside the state sphere.Unsurprisingly, most studies presenting evidence in favor of profit-making LFPSs have been in fact produced or funded by LFPS entrepreneurs or investors (see for instance the controversy raised between Tooley and Longfield 2015] and Day Ashley et al.'s 2015] DFID-Commissioned 'Rigorous Literature Review' on the role and impact of private schools in developing countries).The idea of profit as a necessary factor included in the LFPS definition intensifies only one dimension of the 'entrepreneurial spirit' that may lead an individual or a group of individuals to create a school (Walford, 2011).It automatically leaves out a whole range of undertakings form the social economy sphere such as NGOs, community-run schools or even faithbased organizations that seek to increase access for disadvantaged children (see Nishimuko 2009] for the case of NGOs and religious schools in Sierra Leone, and Blum 2009] for the case of NGOs in India).
Therefore, we consider that the notion of 'low-fee private school' can over-simplify and even blind the view of a phenomenon that is actually much broader and multifaceted (Junemann & Ball, 2015).It needs to be named and considered in such a way that reflects that nature.If these undertakings are coined in more general terms and analyzed at a local scale, then some common structural elements found among different contexts are manifested.A comparison between low-fee private schools in Buenos Aires and low-fee private schools in Asia and Africa may appear hard to accept.However, as Walford (2011, p. 402) puts it regarding his own comparative study dealing with low-fee private Christian schools in England: "I recognize that this comparison may seem somewhat bizarre, however there are occasions when the results of such comparisons can challenge our perceptions and modify our understandings of situations in other countries."In our case, we have identified a potential continuity between both contexts: Buenos Aires' long-standing public-private partnership in education may reflect a possible future for countries that decide to engage in the integration of LFPS into public-private partnerships.
There are various reasons why these schools have been set up in Buenos Aires, a feature shared by the ventures in Asia and Africa where private providers, NGOs, philanthropic organizations, communities and faith-based organizations shape educational systems meeting excess or differentiated demands (Heyneman & Stern, 2013;Rose, 2009).Thereby, personal or institutional reasons include aspects related to religion, philanthropy, commerce, and community engagement.If we narrow these reasons to commercial matters, we would be losing sight of a whole range of solutions that communities build to face the need for education.We would also overlook the diversity of methods that states can promote to govern the system.In general, the development of these schools was or is associated with some kind of state supply deficit, and thus these processes have been characterized as occurring by default.In cases like India or Nigeria, this is evident because we observe these processes in real time as governments explore different alternatives to reach the Education For All targets.In Buenos Aires, with full school attendance rates, we arrive at the same conclusion visualizing a counterfactual scenario where the private sector did not exist and 50% of school enrollment would become 'excess demand' straightaway.This counterfactual scenario becomes even more challenging in vulnerable areas of the city, where private schools hold high enrollment rates and public schools gather in absolute terms twice the number of students when compared to other communes.
Considering this somewhat arbitrary comparison between both contexts, we can see that while in India or Nigeria a great number of these schools are not officially acknowledged, or operate under an irregular system as described by Srivastava (2008), in the case of Buenos Aires, all of them are included within the official system and are thus subject to government regulation.However, it is interesting to note that this acknowledgement was the result of a long process and, in fact, private schools were authorized to grant official academic certificates only from 1969 onwards (Perazza, 2011).The official acknowledgement and the subsidy allocation system were the response of the Argentine state to the challenge regarding the development and administration of a national education system.This public-private partnership solution was chosen then among other possible schemes.
In contrast, national education systems are currently under construction in most countries in Asia and Africa where the LFPS phenomenon has been studied.Primary school net attendance ratios barely reach 65% in West and Central Africa and 80% in South Asia (UNICEF, 2014).In an attempt to achieve the articulation of those educational systems and ensure universal coverage in primary education there are numerous initiatives that promote the integration of LFPSs in publicprivate partnership frameworks of varied and complex nature that need thorough analysis (Akyeampong, 2009;Srivastava, 2010).The long-standing Argentine PPP framework may yield some lessons.It is quite clear that these diverse private undertakings originate in response to the nature of the existing state sector and, as noted, changes in the shape of the state sector determine changes in the balance point of the provision structure.As Woodhead, Frost and James (2013) note, strong government regulation is essential in any public-private partnership arrangement meant to be comprehensive.Reforms in the state sector are also necessary in order to avoid ghettoization of public schools, as evidenced by Vasavi (2003) in India, and Gasparini et al. (2011) in Argentina.In both contexts the questions raised by Härmä (2013) still remain, regarding the impact of private education for vulnerable sectors in terms of equity.Schemes that oblige deprived families to divert part of their income in order to pay for their children's basic education seem inconsistent within a sustainable development program.

Figure 1 .
Figure 1.Private Schools in Communes 4, 7, 8 and 9 according to Type of Entity (Legal Form) -Year 2014 Note: Own preparation based on data from the General Department of Private Education of the Autonomous City of Buenos Aires (DGEGP)

Figure 2 .
Figure 2. Histogram of Monthly Private Schools Fees in Communes 4, 7, 8 and 9 -Year 2014 Note: Own preparation based on data from the General Department of Private Education of the Autonomous City of Buenos Aires (DGEGP) Fees expressed in USD at the 'free' exchange rate.

Table 1
Percentage of Students in Primary Schools by Commune and Type of School.City of Buenos Aires -Year 2012

Table 2
Percentage of Students in Primary Schools by Income Quintile and Type of School.City of Buenos Aires -Year 2012 Note: Annual Household Survey of the Autonomous City of Buenos Aires, and Narodowski, Rozada,

Table 3
Schools, Sections, Enrollment and Indicators per Commune.Primary Level.City of Buenos Aires -Year 2011

Table 4
Percentage of Private Schools in Communes 4, 7, 8 and 9 according to Received State Funding -Year 2014 Own preparation based on data from the General Department of Private Education of the Autonomous City of Buenos Aires (DGEGP) Note:

Table 5
Percentage of Private Schools in Communes 4, 7, 8 and 9 according to State Funding Granted and Religion -Year 2014 Own preparation based on data from the General Department of Private Education of the Autonomous City of Buenos Aires (DGEGP) Note: