An empirical test of a new theory of economic…
An empirical test of a new theory of economic reform using Indonesia as a case study 1988 to 2005
Abstract
Debate regarding key factors determining the success or failure of policies of liberalization and privatization, illustrates the need for a concise theoretical foundation to guide decision makers as to how, when, and where to apply policies that change underlying economic structures. Theories of economic growth have failed to identify key factors underlying economic growth, as well as develop adequate yardsticks of development. Moreover often economic growth is taken as a measure of success without considering the underlying social development and distribution of wealth. A new theory is expounded herein and tested using Indonesian time series data. The results of the empirical tests of this theory reported herein add another dimension to the debate on emerging markets. One factor appears vital to the success of such policies. This is the education level of females. It is the only factor that appears consistently correlated with successful achievement of reform packages, measured by two new criterion of economic and social development.
1. Aims and objectives.
This research has one purpose – to conduct an empirical test of a new theory of Economic Reform. The reason for using Indonesia as a case study is due to its suitability as a country going through a reform process after a period of financial crises and regulatory failure.
Since the fall of the Berlin Wall and the opening up of Communist nations, economists have promoted new ownership structures as a solution to the maximization of welfare and a way of introducing market forces into command economies. They have advocated privatization in particular as the optimal means of achieving an increase in economic growth and lifting per capital income. Privatization is not the only means of reducing state ownership and control of enterprises. Economists have also touted public private partnerships and private finance initiatives as means of promoting economic and social development. However, regulators are still intensely debating whether privatization and other ownership structures, which are not totally dependent on state ownership and control, achieve such goals or in fact cause a deterioration in welfare levels from those existing under a command economy. For instance Stiglitz (2002) claimed that the IMF advocated such policies without considering factors vital to the suitability of such reform programs to a particular economy and society. Reformists did not understand the particular history, the social capital, the political institutions, and how political forces affected political processes (Stiglitz, 1999, p.4) . Academics and other advisers have refuted Stiglitz’s trenchant criticism of development strategies, which rely largely on changing ownership structures, claiming success for such policies in China and Poland (Dabrowski, Gomulka and Rostowski, 2001) . Dabrowski et al (2001) point out that Stiglitz ignores the principal reasons for failure of these policies in Russia, and fails to distinguish why these policies succeeded elsewhere.
The arguments of these authors (Dabrowski et al, 2001) illustrate some important flaws in contemporary thinking regarding the measurement and assessment of development yardsticks, which have implications for advocacy of an ‘optimal’ model for reforming and restructuring an economy. They advocate policies to achieve success in reform by concentrating on improving the institutional, legal, and economic conditions for rapid and sustainable growth. Hence success should be measured by the increase in output deriving from the private sector from the start of recovery, (Dabrowski et al, 2001, p.297), and not by the overall growth of GNP of the whole economy, as suggested by Stiglitz, as this indicator can be affected during transition by pre-reform crisis conditions. Dabrowski et al (2001) advocate a principle where loss making State Owned Enterprises (SOEs) should be shutdown and not privatized, and the rapid expansion of a new private sector encouraged instead. The welfare costs associated with discontinuing and not privatising SOEs can be regarded as an investment “needed to achieve permanent welfare gains from the better allocation of labour and other resources in the future” (Dabrowski et al, 2001, p.298).
This research attempts to take the arguments of both Stiglitz et al (2001) and Dabrowski et al (2001) and test my own formulation of a new theory of economic reform. This will aid in understanding the principal factors that can constitute barriers to the success of economic reform strategies, which rely on new ownership structures to promote efficiency, and thus raise the production frontier of an economy.
2. A new theory of Economic Reform
Debate regarding key factors determining the success or failure of policies of liberalization and privatization, illustrates the need for a concise theoretical foundation to guide decision makers as to how, when, and where to apply policies that change underlying economic structures. Below I outline such a framework, which is based on the perception of gradations in the process of development. It also argues that the introduction of new ownership structures, market mechanisms, and financing techniques are not necessarily solutions without providing for changes in economic, societal, and legal infrastructures.
The new theory of choice of ownership structures espoused in this paper, thus conceives of a national economy as set of interrelating systems and subsystems. Hence the method of liberalising an industry through changing ownership structures can be described in terms of a matrix, (see Table 1), and a set of equations. This in turn dictates a staged approach to changing ownership structures.
In this theory Economic development (Y1) is defined as sustainable growth. This can be measured at the level of the individual by the increase in a maintainable and stable level of income per capita, and at the corporate or institutional level by the increase in maintainable and stable accumulated earnings per capita. At the country level, improvements in the ratio of external debt and current account balance to Gross Domestic Product (GDP), as well as increases in the level of maintainable and stable GDP per capita, are appropriate measures of success. These definitions avoid the criticism of using inappropriate yardsticks of growth of GDP, as in transition economies, growth may initially be negative as pre-reform crisis conditions impact immediately after reform. This was one of the areas of disagreement between Dabrowski et al, (2001) and Stiglitz (1998a,b, 2001) . Also a policy may appear to successful in the short term using growth in GDP as a yardstick, but it is the maintenance of income per capita, corporate profitability and an increasing debt servicing ability at the national level over a long term that is the best criteria of success.
Social development (Y2) is defined as growth in the equitable distribution of wealth, which can be measured by the dispersion and distribution of per capita income, and participation in institutions. A scale ranking the capacity to participate in the institutional framework of government may be a useful adjunct to wealth distribution measures. However, active participation is in turn is a function of improving the quality of human capital, in terms of education, knowledge and skills, which are vital to social and economic dev. This is likely to provide development strategies that adjust policies regarding ownership structures to the underlying fundamental differences between economies.
|
|
Ownership of Capital – proportion of direct and indirect ownership by individuals vs. institutional/elite vs. State (O) |
Type of Ownership Structure, Methods of Valuation and Timing(T) |
| Stage of Social Development(Y2) |
Regulatory Structure (M)and Industry Structure (C) |
Stage of Managerial Expertise, Organizational Structure and Processes (E) |
| Stage of Economic Development (Y1) | Compliance or stage of legal infrastructure development (L) |
Political System (P) which influences Government Goals (G) and which influences the allocation of Economic Resources (R) |
The type of government will influence government goals (G) and in turn will determine the allocation of resources. The optimum set of government goals, are derived from theories of financial regulation (Sinkey, 1992) . These theories emphasise the pre-eminence of safety of depositors in financial institutions, currency and price stability, industry structures that promote competition, convenience of users of financial services (such as access to a product or service) and general public confidence in the financial system. Such goals are considered necessary to promote a market economy. However, how a government allocates its economic resources can determine whether such goals are achievable. An economy can expand its resource allocation through spending more than it raises is revenue, assisted by direct foreign investment, borrowings, and aid. Hence, government goals can be expressed as,
G = f (S, S, S, C, C) = f (R) = f (D, FDI, K, A) (1)
Where safety (S), stability (S), structure (S), convenience (C) and confidence (C) are a function of how economic resources (R) are allocated and may require deficit spending (D), direct foreign investment (FDI), private capital formation and borrowings, (K) and foreign aid (A) .
The optimum mix of direct and indirect public ownership, where the latter is defined by ownership by pension or mutual funds or employee share ownership plans, can be seen as dependent on simultaneous design of the regulatory model and the industry structure, as well as the stage of development of the legal infrastructure. The first two factors of M and C will both influence, and be dependent on, the stage of social development more than L, which has a closer tie to the stage of economic development. This could be expressed as,
The Optimum Ownership Mix or O = M.C (Y2). L (Y1) (2)
The type of ownership structure, whether a PPP, PFI or a privatization, as well as the timing and valuation method is influenced by the E factor – the stage of managerial expertise, organizational structure and processes – as well as by the P factor. The E factor influences and is dependent on social development while the political system, the P factor, has a closer relationship with the stage of economic development. Another way of expressing this is,
The Optimum Type of Ownership Structure or T= E (Y2). P (Y1) (3)
It is postulated that in the above model Social Development, or Y2, may decrease then increase as M, C and E change. This is due to the initial effects of deregulation of protective measures on an industry, and the time lag before learning effects of change kick in and prudential supervision increases in strength to compensate for the reduction in protective measures. The resulting increase in social development may be at a decreasing rate as decreasing returns to scale of M and C are experienced, as the regulatory model and market structure may become burdensome, clogged and lead to obstacles. This requires input from factors affecting economic development.
The political and legal systems must allow feedback so that as social development occurs, more reliance can be made on self-regulatory market mechanisms. Hence the regulatory model and performance of the market structure will require continual monitoring. Development of the E factor will contribute to more effective prudential supervision through enhanced skill levels both within the regulator, the marketplace and the entities subject to changed ownership.
Economic Development or Y1 may increase at a decreasing rate as efficiencies are realized or could display a Cobb Douglas pattern due to the political system not adapting quickly to sectoral imbalances, particularly in the supply of human capital. Alternatively, it could be due to the failure to adjust the legal infrastructure as the quality of human capital increases, which is a sequencing problem. If economic development is uneven, this could promote political instability, which in turn affects resource allocation. This would be most evident in failure to assess and adjust social development policies according to our allocated priorities of M, C, and E.
The implications of this theory, or choice matrix, are that it is vital to assess the stage of P, E, L, M and C – the political system, human capital, legal infrastructure and the regulatory models governing the financial system and industry – before choosing the ownership structure as well as timing and valuation methods and,
- That the design of M, the regulatory model governing the financial system, is a starting point which must take account of the type of existing, and desired market structure, i.e. C, as well as the caliber of E;
- That if all of the factors are weak at the starting point of economic reform, a careful staged approach should be used. Only a formal deregulation and program of ownership change should be considered. This will involve continual monitoring both by the government undertaking the change, and by external aid agencies such as the IMF and World Bank. Otherwise an entrenched elite will take the place of the state in ownership and control, without the requisite spread of the advantages of moving an economy to a full free market basis.
Figure 1: Interrelationships between factors
The interaction of the variables of P, E, L, M and C is multidimensional, but an attempt is made in Figure 1 to illustrate how an economist could diagnose an economy, and prescribe ownership mechanisms as well as necessary reform of factors to promote economic and social development. The purpose of Figure I is to illustrate that a staged approach to the choice of the development of ownership structures must be the overriding principle dominating an economic reform package. For instance the left hand lower quadrant starts with state ownership and control where all factors are weak as opposed to the full market based system in the top right hand quadrant.
We can see from Figure I that where E is rated highly together with M, C, and L that a full public float with reliance on the market mechanism should promote the optimum outcome. This is what we would expect of an advanced economy with a high level of managerial expertise, organizational structure, processes, regulatory and industry structures and legal development. Or if the industry by its nature is non-contestable, such as telecommunications, a privatization can work with prudential oversight of pricing and market practices. That is the C factor must be improved.
Figure 1 also illustrates that a PFI is the ownership mechanism where all factors are rated at mid point, while a PPP is appropriate where all factors receive a low rating, except for the E factor of managerial expertise, organizational structure and processes, which needs to be injected by the private sector. Indeed this is often the rationale for involvement of the private sector. PPPs can occur in an advanced economy, where due to a long history of state involvement in a particular industry such as education, the industry is one where the regulator does not have the supervisory skills or requisite powers. Also the expertise in modern market practices within the industry may be lacking and the industry may be protected from competition by legislation. Hence it appears as a pocket of state ownership and control within a free market economy. An injection of market expertise from allied industries, which have developed private sector practices, should be sought in a consortium between SOEs and private firms.
Figure 1 hence illustrates that where the E factor is weak, while the other factors have been improved to a moderate level, changing ownership structures is not advisable until either the E factor is improved, or the other factors strengthened.
Such a staged approach requires measurement techniques to assess the state of the input factors. The taxonomy of regulatory models described elsewhere (Currie, 2000) could be used to assess the M factor, while the C factor could be assessed by a Porter style industry analysis. The E factor could be measured by the quantitative methods to assess human capital described in Section 2, while the P factor assessed by stated government goals, as well as the resources devoted to all the other factors, as revealed in the national accounts. The L factor could be assessed not just by compliance with the compendium of standards, but by studies of corruption and independence of the judiciary such as undertaken in Indonesia.
3. The need for empirical testing
he theory outlined above has canvassed some of the essential ingredients in the resolution of the debate as to the merits of changing ownership structures on economic and social development. What is new is a theoretical framework to analyse and measure readiness of an economy for reform, as well as to analyse why past attempts failed. This allows a basis to prescribe an optimum staged approach. Indonesia is a fertile database having suffered a severe financial crises with closure and collapse of its banking system after an attempt to privatize sectors of the economy, and now being under IMF direction to privatize as many industries as possible. Data will relate to the measurement of factors detailed in the theory with an attempt to validate hypotheses embodied in the above three equation.
Analysis of past successes and failures in Indonesia in privatizing industry during the period 1988 – 2003 is to be measured by my unique indices for economic and social development (Y1,Y2 ). The ability to empirical assess the input factors of the model, and then relate them to desired outcomes in terms of my economic and social development indices, may lead to other approaches to economic reform through changing ownership structures. For instance, continued government involvement by requiring a percentage of future profits or gain upon resale when economic and political opinion considers privatization a fire sale solution, has not been attempted. Neither have mechanisms to ensure employees (and customers) share in the benefits conferred by a privatized entity.
A basis for such initial research would be a classification of regulatory models in the industries in Indonesia that did and are still to undergo ownership changes as specified by the Republic of Indonesia in conjunction with the IMF, and assessment of the state of requisite inputs in terms of C, E, L, P as well as specification of G and R, government goals and resource inputs. Monitoring of the performance of the industry and entity at a micro level (which can be attempted using measures of profitability and productivity) is essential as is attempting to statistically derive functional relationships between the input of C, E, L, P and output variables which measure success of policies, such as increases in the proportion of GDP the private sector is contributing, as well as the measures for economic and social development (Y1, Y2) as uniquely defined in my theory.
The key dependent variables of interest are changes in the ownership structure, the timing and valuation and how they relate to the independent variables as specified in equations 1,2 and 3 above. Initially multiple regression analysis will be attempted followed by use of curve fitting software to test formulations of the theory.
5. Methodology and Results
The first part of this research involved attempting to develop a new index of economic development known as @page Section1 {size: 612.0pt 792.0pt; margin: 72.0pt 90.0pt 72.0pt 90.0pt; mso-header-margin: 36.0pt; mso-footer-margin: 36.0pt; mso-paper-source: 0; } P.MsoNormal { MARGIN: 0cm 0cm 0pt; FONT-FAMILY: “Times New Roman”; FONT-SIZE: 12pt; mso-fareast-font-family: “Times New Roman”; mso-style-parent: “”; mso-pagination: widow-orphan; mso-ansi-language: EN-US; mso-fareast-language: EN-US } LI.MsoNormal { MARGIN: 0cm 0cm 0pt; FONT-FAMILY: “Times New Roman”; FONT-SIZE: 12pt; mso-fareast-font-family: “Times New Roman”; mso-style-parent: “”; mso-pagination: widow-orphan; mso-ansi-language: EN-US; mso-fareast-language: EN-US } DIV.MsoNormal { MARGIN: 0cm 0cm 0pt; FONT-FAMILY: “Times New Roman”; FONT-SIZE: 12pt; mso-fareast-font-family: “Times New Roman”; mso-style-parent: “”; mso-pagination: widow-orphan; mso-ansi-language: EN-US; mso-fareast-language: EN-US } DIV.Section1 { page: Section1 } Y1. To do this we regressed growth in GDP per capita for the years 1988-2003 using data from the Asian Development Bank (ADB) and the World Bank against a range of variables, using ANOVA techniques. After six rounds the best result was achieved by using the variables of gross domestic capital formation as a percentage of GDP, gross domestic saving as percentage of GDP, the current account balance as a percentage of GDP, external debt as a percentage of Gross National Income (GNI) and Debt Service as percentage of exports of goods and services. The results are described in Table 2.
| Table 2: SUMMARY OUTPUT (Y1 as growth in GDP per capita) | ||
| P-value | Significance | |
| Intercept | 0.00 | |
| Gross domestic capital formation | 0.03 | S |
| Gross domestic saving | 0.05 | S |
| Current account balance | 0.07 | ? |
| External debt as % of GNI | 0.00 | HS |
| Debt service as % of exports of goods and services | 0.03 | S |
Hence we adopted growth in GDP per capita as a measure of sustainable economic growth. We then attempted to develop a measure of social development – Y2 – by regressing two measures of social development – HDI and GINI/HDI against a range of selected variables including the variables in Table 3 plus daily newspapers produced, immunization proxies, percentage of females in the workforce, infant mortality, land use.
Table 3: SUMMARY OUTPUT (Y2 as GINI/HDI,)
| P-value | Significance | |
| Intercept | 0.28 | |
| School enrollment, tertiary (% gross) – SS1 | 0.00 | HS |
| School enrollment, tertiary, female (% gross) – SS3 | 0.00 | HS |
| Personal computers (per 1,000 people) – SS4 | 0.10 | |
| Maternal Mortality Ratio (per 100,000 live births) – SS5 | 0.34 | |
| Life expectancy at birth, total (years) – SS6 | 0.27 | |
Twelve runs resulted in the percentage of the population enrolled in school and tertiary education, in particular females, having the highest correlation with measures predicting a higher index of human development adjusted for a measure of income equality. Hence this measure – GINI/HDI – was taken as a good proxy of social development or Y2.
Given the difficulties in measuring inputs required for the first part of the matrix – that is the determination of the optimum mix of direct and indirect ownership over the time period specified it was decided to concentrate on the third equation –
The Optimum Type of Ownership Structure or T= E (Y2). P (Y1) (3)
This was done by using Standards and Poor ratings system as a measure of the type of political system which combines measures of democracy with political corruption. The Corruption Index provided by the World Bank only extended back to 1995.
Curve fitting was conducted in several stages.
E as a function of Y2
Data for E, education levels measured by the variable of school enrollment, as function of Y2.
The resulting equation which was the line of best fit was
y=a+blnx/x
where
y = Y2
x = E
This run produced a line of best fit with the highest R2 score, where a =16.691747, b= –24.698196.
As the dependent variable was measured by GINI/HDI the results showed that a high level of inequality was associated with a very skewed or unequal level of income and human development and was a function of very low school enrollment – as the school enrollment increased, so did our measure of social development, albeit as a decreasing rate.
P as a function of Y1
Data for P, a Standards and Poor measure of the political system, was fitted as a function of Y1.
The resulting equation which was the line of best fit was
y=a+b/x2
where
y = Y1
x = P
This run produced a line of best fit with the highest R2 score, where a =8.9270203, b= –506.15478.
The results showed that as the underlying economic development indicator improved so did the political system.
T as a function of E (Y2). P (Y1)
T was run as 3D function of predicted-E and predicted- P.
The resulting equation was
lnz=a+bex/wx+c/x2
where
z = T
x = E-predict
y = P-predict
This run produced a line of best fit with 91.2% accuracy (R2), where a = -12.043073, b= -853.28273 and c=2213.8574.
The results showed that Z or the optimum type of ownership structure was not a function of P but of E, with increasing returns to scale up to a certain point, then decreasing returns to scale. In other words, increasing the education level could move an economy along the scale from the weakest to the strongest form of reduced government involvement in the ownership structure.
T as a function of E_predict and P_predict.
Table 4A and B describes the raw data and the output for where T was run as a function of the (predicted-E x predicted-P) to see the best fit equation.
T = a + [ b ln(E.P) / (E.P) ]
was the resulting equation where
y = T
x = E.P
and a=0.42503439 and b=-2.8746624.
| T E-PREDICT P-PREDICT T E.P |
| 0.02412 15.103759 1.37073 0.02412 20.70318 |
| 0.03462 14.963231 3.1568118 0.03462 47.2361 |
| 0.06937 14.144091 4.840355 0.06937 68.46242 |
| 0.08148 13.599148 6.2465353 0.08148 84.94756 |
| 0.09782 13.420381 6.2504787 0.09782 83.88381 |
| 0.18141 12.227915 6.3490631 0.18141 77.6358 |
| 0.26874 11.989031 6.5106916 0.26874 78.05688 |
| 0.33084 11.81598 6.6389212 0.33084 78.44536 |
| 0.42507 10.624672 6.7201481 0.42507 71.39937 |
| 0.41946 10.621856 6.9084821 0.41946 73.3809 |
| 0.33660 10.619251 7.0715105 0.33660 75.09414 |
| 0.40569 10.61005 7.0924358 0.40569 75.2511 |
| 0.39189 9.7760004 7.1703154 0.39189 70.09701 |
| 0.25966 9.7124232 7.2904807 0.25966 70.80823 |
| 0.23345 9.6172235 7.304147 0.23345 70.24561 |
| 0.27860 9.4687832 8.7287513 0.27860 82.65065 |
The results showed that as the education levels interacted with the political system, there could be decreasing returns to scale of reduced government involvement.
6. Conclusion
Despite limits to this research of one case study, lack of data, the results clearly showed a major input into the economic and social development and hence the stages by which economic reform is introduced are education levels, in particular those of the female population. This has policy implications in terms of the importance of education policies and spending on the most vital ingredient to the success of policies which aim to modernize a nation.
There is an obvious need to extend this study across a number of nations as well as test the other equations in the model. As the database for nations becomes more developed this task will become easier.
References
Abdel-Magid, M.F.,1999, Privatization and Valuation: An Accounting Perspective, paper presented at the Global Finance Annual Conference, April 7-9, Istanbul, Turkey.
Baumol, W.J., Panzar, J.C., R.D. Willig, 1982, Contestable Markets and the Theory of Industry Structure, Harcourt Brace Janovich, San Diego.
Beck, T., Demirgüç-Kunt, A., and Levine, R., (1999), A New Database on Financial Development and Structure, World Bank, June.
Binhadi, 1995, Financial Sector Deregulation Banking Development and Monetary Policy, The Indonesian Experience, 1983-1993, Institut Bankir Indonesia, Jakarta.
Buchanan, J., 1986, Public Principles of Public Debt: A Defence and Restatement, Richard D. Irwin.
Caprio, G., Klingebiel, D., 1999, 2003, Episodes of Systemic and Borderline Financial Crises, Policy Research Working Paper. World Bank, Washington, D.C.
Currie, C.V., 2000, The Optimum Regulatory Model for the next Millennium – lessons from international comparisons and the Australian-Asian experience, in Gup, B., (Ed.), New Financial Architecture for the 21st Century, Quorum/Greenwood Books.
Currie, C.V., 2001, The Value of Privatization: The Case of the State Bank of NSW, Economic Papers, March.
Currie, C.V., 2002, “Coping with Systemic Crises — are there alternatives to the IMF medicine?” Seminar sponsored by the Indonesian Chamber of Commerce and Industry, Kompas newspapers (Jakarta Post) and TV 7 – www.
Culpin, R., 1999, Bank Privatization in Turkey: Issues and Suggestions, paper presented at the Global Finance Annual Conference, April 7-9, Istanbul, Turkey.
Dabrowski, M., Gomulka, S., Rostowski, J., 2001, Whence Reform? A Critique of the Stiglitz Perspective, The Journal of Policy Reform, 4, (4).
Gauche, J., 1998, Binary Economic Modes for the Privatization of Public Assets, The Journal of Socio-Economics. 27.
Grabosky, P. and Braithwaite, J., 1986, Of Manners Gentle – Enforcement Strategies of Australia Business Regulatory Agencies, Oxford University Press, Melbourne.
Haggis, I., 1997, Management Accounting and the Polish Privatization Programme, Management Accounting, London, November, 75, (10), pp. 14-16.
Helik, J. 1997, Privatization: The Potential and the Pitfalls, Special Report, Canadian Investment Review, Spring, pp. 23 – 25.
Kelso, L.O., Kelso, P., 1991, Democracy and Economic Power: Extending the ESOP Revolution through binary Economics, University Press of America, Lanham.
Mitnick, B.M., 1980, The Political Economy of Regulation, Columbia University Press, N.Y.
Montanheiro, L., 2001, The Importance of Private Finance Initiatives in Public-Private Partnerhsips, in Montanheiro, L., Berger, S., Skomsoy, G., (Eds.), op. cit.
Moore, D., 1989, Debt: What Should B Done, Institute of Public Affairs, ACT.
Radygin, J., 1997, The 11th Advisory Group on Privatization’s Session on Banks and Privatization, OECD, September.
Sinkey Jr, J.F., 1992, Commercial Bank Financial Management, Maxwell MacMillan.
Stiglitz, J., 1998a, Must Financial Crises Be This Frequent and This Painful?, McKay Lecture, Pittsburgh, Pennsylvania, September 23.
Stiglitz, J.,1998b, Towards a New Paradigm for Development: Strategies, Policies, and processes, Prebisch Lecture, UNCTAD, Geneva, October 19.
Stiglitz, J., 1999, Whither Reform: Ten Years of the Transition, Annual Bank Conference on Development Economics, Washington, D.C., World Bank, April 28-30.
Stiglitz, J.,2002, Globalization and its Discontents, W. Norton, N.Y.
Stiglitz, J., Ellerman, D., 2001, Not Poles Apart: ‘Whither Reform?’ and ‘Whence Reform?’, The Journal of Policy Reform, 4, (4).
Stiglitz, J., Weiss, A., 1981, Credit Rationing in Markets with Incomplete Information, American Economic Review, 7.
