IV. The impact
The hoped-for benefits from decentralization encompass improved efficiency, equity and economic and political participation. Decentralization may alter the allocation of expenditure towards a greater focus on social priorities and local infrastructural projects, promoting human development and encouraging economic participation. Efficiency may be improved as decisions are made closer to where the projects are located, with reduced bureaucracies. Decentralization can contribute to greater equity in the distribution of expenditure, though this also depends on central government measures to redistribute resources as well as the nature of local government. Finally, decentralization should increase political participation of people in the decisions which affect their lives.
This section will assess some of the main effects of Chilean decentralization. There are, however, problems in obtaining rigorous results:
First, Chilean decentralization has generally not involved devolution, as we have just seen, but deconcentration and delegation, under which the national government retained considerable powers with choices at the local level severely constrained. Under Pinochet, moreover, the decisions at the local level were made by central appointees and did not reflect local preferences. Decentralization in such a 'command/control' situation may be quite different from what would obtain if full local input occurred.
Second, there were other simultaneous changes affecting Chile. Two were particularly important -- a general squeeze on government expenditures, which meant that the decentralization of government spending on education and health was accompanied by falling real resources per capita, and a strong move to promote privatization of government services, especially education, which affected the quality of (and resources available for) publicly provided services.
Third, statistics are lacking on municipal expenditure patterns, except in aggregate terms. Empirical analysis must rely largely on regional data and some more detailed sectoral work, thereby limiting the depth of our findings.
A. Human Development Ratios.
For the country as a whole, the proportion of public resources going to "priority" areas in the social sectors15 rose slightly between 1977 and 1988 (Table 11). A breakdown of the various human development ratios shows that the expenditure ratio fell over the period, the social allocation ratio was broadly constant, 16 and the priority ratio rose significantly. The proportion of national income going to human development expenditures thus rose between 1977 and 1988, due to a rising priority ratio which more than compensated for some decline in the expenditure and social allocation ratios, especially the latter.
The education sector showed a sharp fall in the social allocation ratio and a big rise in the priority ratio. These changes reflect the government's withdrawal from tertiary education. A more detailed breakdown of educational expenditure (Table 13) shows a particularly marked rise in pre-primary education, a sharp fall in "other" education (which includes adult education), while the primary/secondary ratio remained constant. In health, there was a slight fall in the social allocation ratio accompanied by a marked fall in the priority ratio, in sharp contrast to the education sector.
The consolidated central government accounts include transfers to municipalities for basic education and health. Deducting these (see Table 12) reduces the expenditure ratio and the social allocation ratio, and makes their decline over the period more marked. The central government priority ratio now falls to zero since the priority areas have been fully transferred to the municipalities. In contrast, the municipality expenditure ratios and social allocation ratios rise, reflecting the transferred functions. The municipal social allocation ratio, which had been zero, rises to 32.2%, more than twice as much as the central government's ratio, and the priority ratio for the municipalities is (by assumption and in virtue of the functions transferred) equal to one.
Comparison of central and local preferences is difficult because municipal choices were constrained by restrictions on the use of government transfers. Perhaps the most legitimate method is to compare the 1977 pre-decentralization preferences of the central government with the choices of the municipalities in 1988. A $100 of government expenditure in 1977 generated $6 spent on priority areas in the social sector. In 1988, $100 of expenditures by the municipalities generated $32 spent on priority areas. Thus a switch of expenditure from the national government to the municipalities would increase social priority expenditure by about one-quarter of the amount switched.
It may be necessary to consider the legitimacy of this comparison since it produces such strong results. We can assume that the 1977 pre-decentralization choices were unconstrained choices. But the municipality choices of 1988 were constrained in two ways -- first by function (they could not spend on non-priority areas), and second by the fact that money was specifically transferred to them for health and education, eliminating their freedom of choice. If the first constraint were relaxed, it seems unlikely that the municipalities would begin to spend on non-priority areas. The fact that municipalities spent more on health and education than the amount transferred17 to them suggests that unconstrained choices would not have been very different if the second constraint were relaxed. These considerations, therefore, suggest that, while the magnitude of the effect of a switch in expenditure resources to the municipality may be exaggerated in the above calculations, the direction of the effect is not. However, it is important to recognize that these changes depended critically upon the effectiveness of transmitting policy preferences within an authoritarian state. Preferences were imposed at the national level by fiat, not by using local preferences (i.e. choosing preferences from below) to set policy.
B. Economic Expenditures.
The other major area in which decentralization may affect choices is in decisions about investment and economic infrastructure. In Chile, the municipalities have shown a high rate of investment compared with the central government (Table 14). Looking at investment allocations, as incomes rose the proportion of total expenditure (excluding the transfers for health and education) devoted to investment rose sharply (to as high as 46% in 1988), while the proportion of central government expenditure devoted to investment was below 10%. If the transfers for health and education are included in municipal income, the investment ratio drops, but is still over twice the rate of the central government.
We do not have an aggregate breakdown of municipal investments by sector. The areas covered, according to the (incomplete) accounts for one municipality, include paving, housing, road improvements, "local development" and "youth improvement." The central government allocation of total expenditure for the broadly corresponding areas (economic functions, housing, and culture) was 18% in 1977, falling to 14% in 1988. Thus, despite the severe paucity of data, we can assume that the municipal allocation to infrastructure and housing greatly exceeded that of the central government.
C. Equity and Efficiency.
Since basic education and health changed from a central government to a municipal responsibility in the 1980s, it should be possible to assess the relative efficiency of the two types of government in these areas. There is some evidence for each sector, but in both cases it is difficult to interpret the data with confidence because other changes were occurring simultaneously, as well as the fact that the transfer took place in stages and was only completed in 1988. 18 The transfer to the municipalities of health and education was associated with cost containment and new structures including greater consumer choice and some privatization. Both sectors were seriously underfinanced by the central government, and municipalities were forced to divert investment funds to cover operating expenses. Education has yet to solve its cost and labor problems. Both human and capital expenditures are also quite inequitably distributed.
Education: In this sector, the government had two objectives -- to municipalize and privatize. A government subsidy per pupil was offered (varying with the level of education), available to private "subsidized" schools19 as well as municipal schools. The goal was to allow parental choice to determine the school their children attended and to raise standards through increased competition between schools.
Educational functions delegated to the municipalities included:
In the early stages of the changes, the subsidy was sufficiently large to attract a large number of private schools and there was a massive shift of the school population from the municipal schools. By 1986, 31% of the school population was in private-subsidized schools. The shift caused substantial problems for municipal schools since fixed overhead expenses were unaltered and it was difficult to sack teachers in proportion to falling enrollments. Furthermore, teachers in both types of schools had to spend a considerable amount of time trying to attract new pupils (Espinola, 1991). Another major problem for both types of schools was the falling value of the educational subsidy. Public expenditure per student was cut by 32% between the peak year (1982) and 1987. As we have seen, this was partially compensated for by the municipalities spending money from other areas (including investment) on education.
Analysis of expenditure per head on education by region shows that it was distributed regressively, with higher income regions spending more per person on education than poorer regions (Table 15 and Figure 4). Expenditure was not related to need as measured by primary school completion rates. The falling rank-order correlation indicates some reduction in regressiveness of the distribution between 1980 and 1983.
Municipal schools' performance showed diverse trends during the 1980s (Tables 16 and 17). Achievement tests worsened, especially in Spanish, and the differentials between socio-economic groups widened between 1982 and 1988. When differentiated by type of school, large metropolitan schools did best and small rural schools worst. Comparisons between municipal schools and private schools showed that, on average, the performance of the private schools was superior to that of the municipal schools, but the differences were much smaller when controlling for socio-economic status. For the lower socio-economic groups the municipal schools did better. 20 Other negative trends observed in the municipal schools were considerable cuts in teacher salaries and a rising student/teacher ratio (Table 17). A few positive changes can also be seen, including an increased secondary enrollment rate (the primary rate staying broadly constant) and a reduction in the years necessary to complete both primary and secondary school.
More subjective evidence has been collected by interviewing mothers. 21 There was a generally negative view taken of the transfer of schools to the municipality -- the chief complaints being the frequent change of teachers, poor conditions, and low morale. The qualifications and standards of teaching were also believed to have fallen. 22 More objective evidence suggests that the mothers were probably right about what was happening in the schools, but not necessarily about the causes. The combined effects of reduced resources, the way subsidies were determined, 23 and the immediate consequences of children (and teachers) shifting to private schools all played a role.
Health: The story in health is, to a certain extent, similar to that in education. The municipalization was accompanied by a cut in real resources and also by the introduction of some user charges, with exemptions for mothers, children and the very poor. Public expenditure on health fell from 3.25% of GDP in 1981/2 to 2.4% in 198824 -- partly due to the privatization of contributions and services among the upper income groups. However, the municipalities contributed their 'own' resources to help offset the decline in central expenditures. As noted earlier, the proportion of total health expenditure going to hospitals increased and the share of clinics declined.
Additional data for resource availability in the public health system showed negative trends during the 1980s (Table 18), the one big exception being the number of clinics and health posts. This considerable expansion made health care more accessible to patients. The picture for the use of health services is mixed (Table 19). Consultations for children under 15 increased and there was a rise in the professionally attended births to almost 100%, but consultations among adults fell and the immunization rate deteriorated. Health outcomes were generally positive -- especially among babies and mothers -- with continued (and marked) improvements in infant mortality rates, low birth weight percentage and maternal mortality. Despite these improvements, life expectancy at birth increased only slowly, implying minimal health improvements in the adult (non-maternal) population (Table 19).
Given the resource constraints, these outcomes suggest that municipalities dealt quite efficiently with their new tasks. Two specific studies were carried out on the effects of municipalization in the health sector. One study25 on municipalities with different income levels in the SMR concluded that:
A survey of how people viewed the transfer of health services to the municipalities, similar to the one carried out for education, reported extensive negative comments about aspects of the health clinics, including waiting time and treatment of patients by non-medical staff. 28 But the study did not have a time perspective, making it impossible to tell if these complaints (which are echoed world-wide) indicate any deterioration after the transfer.
The substantial improvement in maternal and child health over these years, despite falling real revenues received from the Ministry of Health, suggests that the transfer of the functions to the municipalities did not reduce efficiency in this sector (and probably increased it). The evidence of distributional disparities in incomes and achievements between municipalities (also observed in the education sector) indicates that special redistributive action -- both in terms of finance and training -- may be needed to counter this tendency.
Investigations of thirteen poor municipalities in the Santiago Metropolitan Region in 1986-7 indicated considerable variation in the overall effectiveness of the municipalities as development agents, ranging from those which were dynamic and innovative to those which were largely passive and responsive. 29 A key difference lay in the character of the mayor: The mayors visualize their role and put it into practice in diverse ways. We met mayors who were energetic and active, who initiated activities in their municipality, and mayors who were passive with little initiative, where local government was almost no government. 30 Another important difference was in the existence of an "equipo de trabajo" -- a workteam -- consisting of well qualified heads of different departments, who met regularly, together with the mayor, to discuss and plan. In the majority of municipalities studied, Raczynski and Serrano found little evidence of initiative. One mayor31 attributed this to the fact that the Common Municipal Fund and the Regional Fund transfer systems under Pinochet provided little incentive to press for increased local capacity. This, coupled with the restrictions on spending and revenue collection, led to municipal governments that were quite passive, fulfilling their functions reactively. In a few cases, the local authorities were active, planning new initiatives, consolidating and developing local capacity. The best examples encountered revealed all the virtues hoped for from decentralization:
They work well -- professionals, with experience, who are immersed in local problems, the doctor from the clinic, teachers, the local police, and municipal officers. They are well managed, with active and participatory meetings. They have beautiful projects. [Comment by someone interviewed on the functioning of local poverty committees -- Raczynski and Serrano, 1988, p. 51]
The diverse sources of funds accessible to the municipalities beyond their own resources, including various special funds, sectoral ministry funds, and NGO funds means that local initiatives can be important in attracting additional resources. The new structures offer considerable potential for an active development role, but the municipalities need more resources, both financial and human (including training) to realize this potential. 32 The training aspect is particularly important, as municipal governments do not have experience in either local development or tax assessment, billing, and collection. Any increase in financial resources would also require new personnel and extensive training to increase local capacity to deal with these issues.
D. The Distribution of Public Investment
Decentralization under Pinochet also had effects on the distribution of public investment. Aggregate investment by municipalities rose very fast, 1977-81, reflecting the increased resource availability arising from changes in municipalities' financial powers; it rose further to 1988, and then fell back to the 1981 level by 1990 (Table 14). The consolidated government accounts show investment moving in the opposite direction falling by a third, 1977-81, and subsequently rising quite rapidly to 1988 and then, like municipal investment, falling back. Data for FNDR show a fall 1980-86 of 3.8% p.a.
A comparison of the distribution of investment financed by the municipalities, that financed by the regional Fund (FNDR) and that carried out directly by the central government (Table 21 and Figures 6-8) shows that none are progressive, and most are regressive, being positively related to income per capita of the region. The most regressive distribution is shown by aggregate municipalities' investment, followed by the central government. The FNDR is neutrally distributed with respect to income but shows a regressive distribution if we take nutrition as an indicator; regions with larger populations receive more from the FNDR than less populous ones. In the case of the municipalities and central government, the regressive tendencies appear to diminish over time, but the FNDR moves in the opposite direction.
The data thus shows that taken together government distribution of investment in Chile in the 1980s was regressive as between regions. From this perspective there is little to choose between the municipalities and the central government. However, the deliberate attempt to make the regional fund progressive, while not successful, did appear to prevent the marked regressive tendencies to be found in the other types of finance.
E. Social and Political Participation.
The nature of decentralization under Pinochet was non-democratic, as municipal elections were abolished. But some "participatory" structures were adopted, including the role of the "Juntas de Vecinos," which were organized into a general union representing all Juntas in the municipality, and the advisory councils (CODECOs) of the municipality. Leaders of the Juntas de Vecinos were appointed by the mayor, and were used as a mechanism for passing information down from the center, rather than the reverse, i.e. it is generally agreed that they were rarely active participatory organisms. The CODECOs consisted of appointed representatives of various local interests to advise on local programs and projects. There has been considerable variation in the functioning of these councils: some rarely met and did little. Others, especially in richer municipalities with a stronger business presence, were more active. But their participatory activities were limited by their advisory functions. Informal associations developed over the 1980s, e.g. mother centers, sporting clubs, housing associations, which were the most active in representing their members' interests. 33
As noted earlier, the structure of government under Pinochet was essentially top-down, in which local governments were instructed to enhance efficiency. Naturally, this limited the role of social and political as well as economic participation. As the World Bank observed: "the Chilean model is structured to amplify messages downward and suppress those from below." 34
F. Macro-balances.
One argument widely advanced against decentralization is that it will reduce macro-economic budgetary control. Hence macro concerns are among the factors determining Chile's choice of decentralization strategy. There seem to be two aspects to the argument. First, while the central government may have control over the central government budget, deficits at the local level can undermine it; and second, even in the absence of macro deficits, the size of the state may be increased by decentralization.
The validity of both types of argument depends on the nature of the decentralization adopted. In the Chilean case, during the Pinochet era there were strict constraints, such as inflexible wage scales for local government employees and an enforced balanced budget law, preventing either of these possibilities from becoming reality. Pinochet's concern with controlling municipal sources of debt and uncertainty prohibited municipalities from running deficits, while the regional governments were denied independent budgets.
In practice, as noted earlier, the situation in education and health did cause deficits in these sectors among the municipalities. To a large extent, the municipalities themselves dealt with the problem by "squeezing" other expenditures in order to transfer resources into these sectors. Occasionally, they appealed to the national government, which dealt with each municipality, using strict financial conditionality to discourage future deficits. In general, deficits were rare; investment was strictly controlled because both regional authorities and the ministry of interior had veto power over major projects. Problems arose, however, because functions were transferred to the municipalities without adequate finance or tax powers.
Under the new government, fiscal pressures are likely to develop at
the local level once municipal and regional levels become more democratic
and have a political base from which they can press for more powers, including
those of raising local taxes and borrowing. The provision of more freedom
to vary local taxation involves the possibility of an increased proportion
of income going into taxes and public expenditure, unless offset by reductions
at the central level. This action would reflect local preferences and therefore
does not present a problem. The issue of local borrowing raises potential
problems for macro-balance, which can be dealt with either by firmly prohibiting
local borrowing, or by such strict central supervision of the micro level
that decentralization objectives could be threatened.
V. Conclusions
We have shown that there was a significant increase in decentralization in Chile over the last fifteen years, beginning with the enhanced delegation of central powers. More recently, decentralization has been transformed into constrained devolution. However, by international standards, the extent of decentralization in Chile is still quite low, and local governments are subject to substantial constraints. Pinochet-style decentralization was not participatory, but this is slowly being corrected under the current democratic government.
Assessment of the impact of decentralization over these years shows that local government priorities are directed to a substantially greater extent towards social goals and economic infrastructure than are central government priorities. Ceteris paribus, a switch in resources towards local governments would therefore be likely to achieve an improvement in resource allocation.
Severe resource constraints and other changes introduced at the same time, as well as data problems, have made it difficult to assess the impact of decentralization on efficiency. Some education indicators worsened during the decentralizing period, but the health indicators continued to improve. Given their constraints, it would seem that the municipalities managed at least as well as the central government had. In health, education and more generally, there is evidence of diverse performance among different municipalities, with richer/larger municipalities doing better than others, though this is probably due to their higher per capita allocation for social investment. Decentralization seems to give rise to the potential for enhanced diversity in performance, and consequently more need for offsetting action to correct disparities. Neither the special national nor the municipal investment funds have been redistributional; indeed total investment remains highly correlated with regional income. The variance in regional income has actually increased by over 2% during decentralization and this, along with local tax base inequalities and an inegalitarian investment pattern, are major challenges to be faced by the new Frei government. We are not claiming that decentralization has caused a worsening of inter-regional distribution, only that it has not managed to overcome the basically inequitable flow of investment funds.
The extent of decentralization in Chile is limited by severe restraints on the financial resources of the municipalities. Their own resources could be increased by increasing their tax powers; their control over resources could be increased by providing block grants from the center rather than a series of special funds, all hedged with a variety of restrictions and controls. However, distribution, training, and welfare issues must also be simultaneously confronted if such additional economic freedom is to be extended to the municipalities.
The new emphasis on the regional level in Chile could represent an opportunity for increased decentralization by strengthening a level of government which could exploit economies of scale while maintaining regional control over priorities. But it might also be a step backward by reducing the power of the municipalities, while the center maintains control over the regional governments. It is too soon to predict the outcome -- indeed much will depend on the new political pressures that emerge at both the local and national levels. But it is our impression that regionalization vintage 1993, i.e. without direct elections, probably represents a regressive step. Overall, though, the new democratic structure of government has represented a substantial step forward from the Pinochet era, as constrained delegation has been replaced by constrained devolution.
Democratization will presumably generate significant continuing pressures
to adjust the system further. Fiscal pressures to increase local spending,
formerly repressed by the Pinochet regime, and the reintroduction of politics
at the local and possibly the regional level, will predictably force the
system to make further adjustments.
Bulnes, L. et al., 1988, La Regionalizacion Juridica de Chile (Santiago, Chile).
Campbell, T., 1990, Local Government in LAC: National Strategies and Local Response in Planning, Spending, and Management: Decentralization in Chile, Country Report (Washington D.C.: World Bank).
Castaneda, T., 1989, "Innovative Social Policies for Reducing Poverty: Chile in the 1980s," mimeo (Washington D.C.: World Bank).
CEPAL, Statistical Yearbook for Latin America, 1991.
Country Development Report: Chile, 1990, (Washington D.C.: World Bank).
Duran, L., N. Herrera and G. Martelli, 1992, "El Sistema Financiero Municipal Chileno: Problemas, Desafios y Propuestas", (Santiago: FAUS).
Espinola, V., 1991, Descentralizacion del Sistema Escolar en Chile (Santiago: Centro de Investigacion y Desarrollo de La Educacion).
El Fondo Nacional de Desarollo Regional, Santiago, Chile; Central Government Document, 1992.
Government Financial Statistics, 1991, Santiago, Chile.
Human Development Report, 1990-1993, (Washington, D.C.: World Bank).
ILPES, 1992, Manuel de Capacitacion Municipal (ILPES/DEIDESCO/PNUD) (Santiago: ILPES).
IMF Monthly Statistics, various years, (Washington, D.C.: IMF).
Irarrazaval, I., 1992, Coordinator, "Problemas y Propuestas para el Sistema de Financiamiento Municipal", Comision de Estudios sobre el Sistema de Financiamiento Municipal, No. 176, April, (Santiago: Centro de Estudios Publicos).
Lashman, S., 1990, Social Development and Progress in Chile: Achievements and Challenges (Washington, D.C.: World Bank).
Loyola, V. and E. Miranda, 1990, "El Nivel Administrativo de Consultarios de Atencion Primaria Municipalizados", Revista Administracion en Salud, 10 (Santiago: University of Chile).
MIDEPLAN, 1992a, "Avanzando en Equidad, Un proceso de Intergacion al desarrollo: 1990-1992" (Santiago: Ministerio de Planificacion y Cooperacion).
MIDEPLAN Documentos/Regionales, 1992b, Serie Regionalizada de la Inversion Efective 1980-1990.
Ministerio de Hacienda. Revenue and Expenditure Data, Santiago, Chile, 1992.
Miranda, E., 1990, "Descentralization y Privatizacion del Sistema Salud", Estudios Publicos, 39, 5-60.
Panzer, V., 1989, "The Municipalization of Health Services in Chile", Article presented at el Encuentro Anual de Economistas (Santiago, Chile).
Prawda, J., 1992, "Educational Decentralization in Latin America: Lessons Learned", A View from LATHR, 27 (Washington, D.C.: World Bank)
Raczynski, D., C. Serrano and Edgardo Bousquet, 1990, "Escuelos y Educacion Basica. La Mirada de Padres y Apoderados en Penalolen Alto", Apuntes Cieplan, Nov., 95 (Santiago: CIEPLAN).
Raczynski, D., and P. Romaguera, 1992, "Chile: Poverty, Adjustment and Social Policies in the 1980s" Paper prepared for the Conference, "Confronting the Challenges of Poverty and Inequality in Latin America", The Brookings Institution and Inter-American Dialogue, July 16-17, 1992, Washington D.C. (Santiago: Cieplan).
Raczynski, D., C. Serrano and Edgardo Bousquet, 1990, "La Comunidad frente a los Programas y Servicios del esado: la Salud Publica en Penalolen Alto", Apuntes Cieplan, Dec., 96 (Santiago: CIEPLAN).
Raczynski, D. and C. Serrano, 1988, "?Planificacion para el Desarrollo Local? la Experienca en Algunos Municipios de Santiago", Coleccion estudios Cieplan, 24, 37-62.
Raczynski, D. and Cabezas, 1988, Ingresos y Gastos Municipales: Chile (1977-1987) y Grad Santiago (1985/86) Notas Technicas, #121 (Satiago: CIEPLAN).
Raczynski, D. and C. Serrano, 1987, "Administracion y Gestion Local: la Experiencia de Algunos Municipios en Santiago", Coleccion estudios Cieplan, 22, 129-151.
Rondinelli, D. Nellis, J., and Cheema, G.S., 1983, Decentralization in Developing Countries (Washington D.C.: World Bank).
U.N. Statistical Yearbook, various years, (New York: United Nations).
World Bank, 1990a, Chile, Consolidating Economic Growth August (Washington, D.C.)
World Bank, 1990b, "Decentralisation in Chile", A LATIE Regional Study, Mimeo, Nov. (Washington D.C.)
World Bank, 1992a, "Chile, Subnational Government Finance", May (Washington, D.C.)
World Bank, 1992b, Green Book (Santiago: Banco Central de Chile).
World Development Report, 1979, (Washington, D.C.: World Bank).
15 Defined here as basic education and primary health care.
16 Data for the social allocation ratio differ. IMF data show a quite sharp fall between 1981 and 1988 while government sources suggest it remained generally constant. Estimates of priority expenditures as a proportion of GDP for 1988 therefore also differ.
17 In 1990, 31% of health expenditure by the municipalities and 10% of education expenditure was financed by the municipalities out of their own resources (Bulnes et al., 1988).
18 In each case, the process took place in a series of steps. In education, transfers to the municipalities started at the end of 1980, but were interrupted by the financial crisis of 1982 (87% of schools had, however, been transferred by the end of 1982) (Castaneda, 1989). There was a similar timetable in health, although less than 20% of establishments had been transferred by the end of 1982; the process was completed from 1986-88 (Miranda, 1990).
19 Fully private fee-charging schools were not eligible for the subsidies, but the private subsidized schools do seem to have been able to charge some supplementary fees.
20 There is also a quite sharp difference between "old" (pre-1980) private subsidized schools and post-1980 schools, with the former (often missionary schools, or developed for some other educational purpose) outperforming municipal schools and the latter (profit-motivated) doing less well (Espinola, 1991).
21 Raczynski, Serrano and Bousquet interviewed 39 mothers in the municipality of Penalolen in SRM in November 1989. Raczynski et. al., Nov. 1990.
22 Raczynski et al., Nov. 1990, p. 13-15.
23 A recurrent comment in the writings (and interviews) on this issue was that the need to attract pupils led to emphasis on rather superficial matters, such as the name of the school and its appearance, while teachers spent too much time trying to attract new pupils, and not enough on existing ones.
28 Raczynski, Bousquet and Serrano, Dec. 1990.
29 Raczynski and Serrano, 1987, 1988.
30 Raczynski and Serrano, 1987, p. 133.
32 Raczynski and Serrano, 1988. They also point to the problems caused by the mayors being appointed by and responsible to the President as well as to the local community. But this was changed in June 1992.
33 Raczynski and Serrano, 1987. Also World Bank, 1990b: "Local councils which once had an important role in articulating community preferences have, under Pinochet, been nullified as an effective political voice." [Executive Summary, p. 6]