Adelman and Morris Stage Theory OR Social, Political, and Economic Stage Theory:

Irma Adelman and Morris presented their theory of stages of economic growth in their book "Society, Politics and Economic Development" in 1967. They share with Rostow and others that the process of economic development can best be analyzed in terms of stages. They use different techniques in distinguishing these stages. Again, they explain growth within each stage and they identify the factors that determine the transition from one stage to another. These two economists deal with political, social, cultural and economic factors by focusing on the experience of LDCs. Thus, they provide empirical results regarding the stages of economic growth relating to contemporary countries.

Adelman and Morris use a variety of statistical techniques, including factor analysis, discriminate analysis, canonical correlation and analysis of hierarchical interactions. These techniques present different forms of multivariate analysis. However, we use Factor Analysis Approach.

Like Regression Analysis, factor analysis is basically the 'Analysis of Variance Technique'. It decomposes the variance of a variable into several components based on its association with other variables. Factor analysis is primarily helpful to organize and simplify complex statistical data. Factor analysis is close to regression analysis. However, it is clarified that regression analysis may make it possible to identify causality and dependence, while factor analysis may be thought of as identifying only interdependence.

The first step in stage theory is to classify or define the characteristics of a stage and to identify countries with respect to particular stages. With reference to a sample of 74 LDCs, Adelman and Morris obtain quantitative or semi-quantitative data for each of 40 different social, political and economic indicators of development. Some of these development indicators are traditional like per capita GNP, but some which are social and political ones are non-traditional like character of agri. organization, extent of leadership, commitment to economic development, degree of modernization of outlook, extent of social mobility and character of basic social organization.

The list of 40 indicators consisting of socio-cultural, political and economic groups is given in table 1.

According to Adelman and Morris, some of the below-mentioned indicators, in turn, based upon two or more sub-indicators. For example, indicator 6, "Extent of Social Mobility" is measured by:

(i) The ratio of the population five to nineteen years of age that is enrolled in primary and secondary schools.

(ii) The importance of the indigenous middle class.

(iii) The presence or absence of prohibitive cultural or ethnic barriers to upward social mobility.

Table 1:

Indicators of Social, Political, and Economic structure utilized by Adelman and Morris:

After determining the indicators and sub-indicators, Adelman and Morris assign each of the 74 LDCs a letter grade with respect to each indicator. Finally, the letter grade scales for these indicators, many of which are only qualitative, are converted to a numerical scale. The resulting "ordinal" scores are the basic data for their application of factor analysis and for their subsequent extensions of that analysis with other tools of multivariate analysis.

Table 2:

Rotated Factor Matrix for Per Capita Gross National Product together with 24 Social and Political Variables.

The first application of the technique is the interaction of social and political indicators in the process of development. For this purpose, all the economic indicators given in the above Table 1, except per capita GNP, are omitted and even this indicator is kept separate from the for factors into which the social and political indicators are clustered. The results appear in Table 2.

Factor 1 (F1 in the table) refers broadly to the extent of social differentiation and integration, i.e., "processes of change in attitudes and institutions associated with the breakdown of traditional social organization".

Factor 2 (F2) is associated with political systems indicating the transition from "centralized authoritarian political forms to specialized political mechanisms capable of representing the varied group interests of a society and of aggregating these interests through participant national political organs".

Factor 3 (F3) relates to leadership, "the strength of industrializing elites relative to traditional elites".

Factor 4 (F4) refers to social and political stability.

Methodology for Determination of Factors (Matrix or Mathematical Approach):

(It is optional for students):

The above mentioned Factor Analysis problem can be expressed in matrix form. As, if we have data for m countries consisting of n indicators such as GNP per capita, level of education, and so on, we can denote the decomposition of the variance of each indicator as:

Formula:

Where x is a column vector of n indicators, A is a vector of 1 x n, B is a matrix of n x q, f is a vector of order q x 1, and U is a vector of order n x 1. In matrix form for m countries, we write:

X = A + BF + U

Where X is an m x n matrix with elements the observable indicators for each country. The elements of the factor F are the latent variables, the factors. B consists of coefficients of these factors, called factor loadings.

The major aim of factor analysis is to determine the factor loadings, i.e., the coefficients that relate the observed variables to the common factor. Factor loadings play the same role in factor analysis as regression coefficients in regression analysis. The squared factor loadings represent the relative contribution of each factor to the standardized variance of each indicator, i.e., xi. If a given factor, ft appears only in a subset of the elements of X, it is called a group factor. It is possible, however, that a factor fi appears in all the elements of X. Then it is called a common factor, and the commonality for each variable is represented by the sum of squares of its factor loadings. Commonality indicates the extent to which the common factors account for the total unit variance of the variable xi. It is akin to the coefficient of multiple determination in regression analysis, the R².

the other indicators in any single factor, but was included in the factor analysis, and thus there are factor loadings for it with regard to each of the four factors. The factor loadings for the per capita GNP are - 0.73 for F1, 0.31 for F2, -0.26 for F3 , and -0.03 for F4.

Since the squares of factor loading indicate the percent of the variance in the variable associated with each of the factors, Adelman and Morris claim that 53% of inter-country variation in per capita GNP in 1961 is explained by F1, and additional 10% by F2, another 7% by F3 and about 1/10^th of 1% by F4. The sum of the squared factor loadings in the case of GNP per capita 70% is the commonality of each indicator, and it represents the proportion of the total variance that is explained by the four factors taken together. The finding that 70% of the variance in GNP per capita is attributed to socio-political indicators grouped in the four factor loads Adelman and Morris conclude that:

"It is just as reasonable to look at underdevelopment as a social and political phenomenon as it is to analyze it in terms of inter-country differences in economic structure."

A further interpretation of rotated factor loading is given as:

"The loading of -0.73 of GNP per capita for F1 and 0.89 for the size of the traditional agri. sector included in F1 implies that GNP per capita is inversely related to the size of the traditional sector.

The authors further apply factor analysis to the same set of data for the 74 LDCa. Since F1 is the most important factor and constitutes a much broader index of development than the other conventional measures, each country is then scored relative to F1. These factor scores, in turn, are used to divide the sample of countries into three groups, identified as different stages of development, the "lowest intermediate and highest" stages.

Separate factor analyses are then computed for three different sets of countries, i.e., regional sub-samples for Africa, Asia and Latin America, with the same set of indicators. Since, the three different regions correspond at least roughly to different stages of development, the authors find the characterization of the factors varies from stage to stage. Adelman and Morris came to the conclusion that the social factors determining inter-group differences in GNP per capita are the lowest group (Africa) and the political factors played the dominant role in explaining such differences at later stages.

In an other application of factor analysis, Adelman and Morris (1967) investigate the interactions of social, political and economic factors using all the indicators in Table 1. This analysis is applied separately to each of the three different stages identified on the basis of the country scores with respect to F1. In these cases, the rate of growth (instead of per capita GNP) is the dependent variable. The results remained more or less the same. They came to the conclusion that the countries at the lowest end of the socio-economic scale, the growth process required to have both social and economic transformation. For the countries at the intermediate level of development, the statistical results are inconclusive. However, it were the economic factors which govern the process of industrialization. Finally, the political pre-conditions for development are important in countries at the high end of socio-economic scale. The crucial correlates of economic performance in these countries are the effectiveness of economic institutions and the extent of national mobilization for development.

Criticism/Demerits:

middle class" may be futile exercise as the case with defining Rostow's "Pre-Newtonian Science and Technology". Again, attaching economic development with social and political variables operationally is imaginative.

(ii) Indicators and their Relationships: The critics are of the view that some indicators are based, at least partially, on the same sub-indicators and thus may have introduced some spurious correlation among the variables. This is specially true with regard to 12 social indicators associated with factor 1 whose results have been given in Table 2.

(iii) Ordinal Measurement of Indicators: According to Brookins, a set of problems arises as a result of the ordinal measurement of the indicators included in the analysis. He is of the view that ordinal variables are inappropriate for deriving' elasticities and 'multipliers'.

(iv) Assumptions are Arbitrary: It is objected that many of the assumptions made at various points in the analysis are arbitrary, and some are quite unjustified. This also looks quite strange when Adelman and Morris did not include any economic indicator (except GNP per capita) in their first formulation of the stage theory. However, in their next formulation (short run analysis) these indicators were included. It also means that the theorists are well authorized to include or exclude the indicators in the analysis.

(v) Beauty of Analysis: Despite certain shortcomings the beauty of the analysis lies in the fact that they have classified complex data and by reading the correlations they have formulated specific hypotheses. On such hypotheses would be that the four factors (the extent of social differentiation and integration, the political transformation from authoritarian regimes to representative govts., the quality of leadership, and the extent of social and political stability) account for the variance observed in a number of socio-cultural and political indicators from a large sample of LDCs. Another hypothesis would be that the relative importance of the factors varies from stage to stage, with social factors being more important in the early stages, political and economic factors in the later stage.

Relevant Articles: