politics, development and social policy: a cross-national analysis

16
European Journal of Political Research 4 (1976) 361-376 361 @ Elsevier Scientific Publishing Company, Amsterdam -~ Printed in The Netherlands POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS GHULAM M. HANIFF St. Cloud State Univerity, St. Cloud, Minnesota, U3.A ABSTRACT Social policies, of the type examined in this study, are greatly shaped by a nation’s level of socioeconomic development rather than by political conditions. The socioeconomic variables that best account for policy differentials among nation- states are wealth and literacy. On the whole, wealth is the best predictor of social policy output. Ceteris paribus, political variables which seem to be important for social policy are political competition and political stability, despite the generally weaker relations between political system characteristics and policy outputs. Generally, it could be predicted with a high degree of probability that nation- states which are stable and which enjoy a high per capita income would also have a comprehensive system of social welfare measures. It has been an axiom of modern political life that nation-states exist to provide for the welfare of their citizens. Despite avowedly professed goals of social betterment, national variations in social policy exist, and the conditions of wide disparity among nations are thought to be social- ly determined. Though differences in national political structure and socioeconomic development have been accepted, very little systematic effort has been made to understand social policy outputs in relation to these national features. This paper is devoted to an examination of social policy outputs, in a cross-national context, by uncovering the political and socioeconomic correlates of policy outputs. The central proposition of this analysis is that the level of social policy is a function of the level of socioeconomic development of the nation-state, affected very little, if at all: by its political complexion. Social policy outputs, as represented by the expenditure of funds, can be correlated with political and socioeconomic variables in order to show and explain the factors associated with policy outcomes [ 11. This technique has been used quite successfully in explaining the structure of policy outputs among American states. In the context of the Ameri-

Upload: ghulam-m-haniff

Post on 19-Jul-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

European Journal of Political Research 4 (1976) 361-376 361 @ Elsevier Scientific Publishing Company, Amsterdam -~ Printed in The Netherlands

POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

GHULAM M. HANIFF

St. Cloud State Univerity, St . Cloud, Minnesota, U3.A

ABSTRACT

Social policies, of the type examined in this study, are greatly shaped by a nation’s level of socioeconomic development rather than by political conditions. The socioeconomic variables that best account for policy differentials among nation- states are wealth and literacy. On the whole, wealth is the best predictor of social policy output. Ceteris paribus, political variables which seem to be important for social policy are political competition and political stability, despite the generally weaker relations between political system characteristics and policy outputs.

Generally, it could be predicted with a high degree of probability that nation- states which are stable and which enjoy a high per capita income would also have a comprehensive system of social welfare measures.

It has been an axiom of modern political life that nation-states exist to provide for the welfare of their citizens. Despite avowedly professed goals of social betterment, national variations in social policy exist, and the conditions of wide disparity among nations are thought to be social- ly determined. Though differences in national political structure and socioeconomic development have been accepted, very little systematic effort has been made to understand social policy outputs in relation to these national features. This paper is devoted to an examination of social policy outputs, in a cross-national context, by uncovering the political and socioeconomic correlates of policy outputs. The central proposition of this analysis is that the level of social policy is a function of the level of socioeconomic development of the nation-state, affected very little, if at all: by its political complexion.

Social policy outputs, as represented by the expenditure of funds, can be correlated with political and socioeconomic variables in order to show and explain the factors associated with policy outcomes [ 11. This technique has been used quite successfully in explaining the structure of policy outputs among American states. In the context of the Ameri-

Page 2: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

362

can state politics one group of scholars have emphasized the primacy of political conditions as the determinant of social policy, and another has articulated the centrality of socioeconomic forces for an explanation of variations in policy outputs [2] . The present study attempts to ex- amine policy variables by relating them to political and socioeconomic variables for a cross-section of 125 sovereign nation-states now com- prising the world political community. In this way it would be possible to assess the relative importance of political and socioeconomic features of nation-states for social policy outputs, enabling one either to accept or reject the central thesis of this study.

The analytical scheme adopted for this work conceptualizes public policy as the terminal product of the political system, structured by the underlying social and economic realities [ 31 . The model constructed for the present analysis identifies the input linkages of the policy out- puts, and seeks to systematica!ly interrelate inputs and outputs through statistical operations.

The social policy outputs selected for analysis consist of five impor- tant policy issues; namely, education, social security, social services, public health, and transportation. The basic task is to account for the factors which are associated with expenditures for these activities by identifying their political and socioeconomic correlates.

Discussion

There are considerable variations in policy outputs among nations; some governments are very active in producing decisions to satisfy the needs of their citizens while others are not. Why do governments differ in the levels of policy outputs when citizens everywhere desire similar services for their welfare? How can these differences by accounted-for? Does the organization of political life make much difference in the eventual output of policy products? What roles do environmental fac- tors of the type associated with industrialization, wealth, literacy, modernity and urbanization play in policy outputs? These questions are crucial to an understanding of public policy outputs in a comparative setting and need to be explored for an adequate description and explanation of policy differentials.

Though the range of variations in social policies pursued by nations is enormous, no satisfactory and consistent theory has been developed to account for these differences. At the level of comparative state politics in the American setting where extensive empirical studies have been conducted there are at least two hypotheses to account for differentials

Page 3: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

363

in policy outputs. Early scholars of American state politics have argued that political competition is the key variable in accounting for policy variance. Indeed, V.O. Key, in Southern Politics presents evidence sug- gesting that states with a loose multifactional party system with little tradition of competition are low in the level of liberal social welfare policy, whereas states with a bifactional one party system, with greater competition, are high in the level of liberal social welfare policy. Simi- larly, Duane Lockard, in New England State Politics, found the two party states of New England that were more competitive, also pursued more liberal social policies than one party states. The linkage between political conditions and social policy remained unexamined until the onslaught of the behavioral revolution in political science, which re- directed the analytical focus of politics into inclusion of ecological vari- ables for an understanding of political life. Richard Dawson and James Robinson, through a more extensive study, relying not only upon political variables but socioeconomic ones as well, completely over- turned the Key-Lockard thesis by demonstrating statistically the strong relationship between socioeconomic factors and social welfare policy [41. This relationship persisted even after controlling for political con- ditions. A number of other scholars further developed this theme, and Thomas Dye probably produced the most exhaustive research [ 5 I .

Although considerable work in this area has been conducted in the American setting, the field of comparative national politics has remained virtually untouched by scholars interested in public policy analysis. Among isolated works specifically focusing on cross-national public policy analysis is one by Phillip Cutright, in which the author examines social security programs for a selected list of nations [6]. Cutright found a fairly strong correlation between economic development and his index of social policy. A more recent book-sized monograph by Alexander Groth examines the output side of comparative politics in a more historical perspective. He argues strongly for the primacy of polit- ical conditions as a determinant of policies designed to enhance the welfare of the ordinary citizen [ 7 ] .

It is unlikely that this paper can resolve the dispute between those who hold the view that political conditions are crucial for social policy, and those who argue that socioeconomic forces are instrumental in shaping policy outputs. As a more modest alternative it was decided to explore the two competing hypotheses by examining a whole series of social policies, with a more extensive data base. In order to facilitate this operation so as to search for conditions to account for variations in the empirical phenomena of interest, a model, specifying variables and the direction of their linkages, was constructed.

Page 4: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

3 64

The Model

One way of imposing order on complex social processes is to develop a model which may, at least, begin by identifying the independent and dependent variables [81. In the scheme presented for analysis in this study, social policy constitutes the dependent variable, and its socio- economic and political correlates constitute the independent variables. The dependent variable represents programs on five of the most impor- tant areas of activity of national governments: education, public health, social security, social services and transportation.

The two sets of independent variables, representing both the socio- economic and political forces of the society, are considered to be the inputs influencing the shaping of social policy [9] . In a general way, the societal structure, taken as a whole, crucially affects the allocation of resources, and determines the amount of social effort that can be devoted to social policy [ 101.

As a conceptual framework for analysis, the model helps in the organization of a set of classifications into which the interdependent variables can be categorized, and subsequently, the suggested relation- ships among them verified with data from the empirical world. Of necessity, the model for this study is an overly simple representation of a complex social phenomenon. since, as a heuristic device, it is meant to clarify thinking about relationships that logically ought to exist in the real world. The relationships portrayed between the independent and dependent variables are represented by the diagram in Fig. 1. The solid lines portray the causal linkages among the interdependent variables that might be said to exist in the social universe. The diagram illustrates inter-connections at one point in time, although the dynamic character of the social process is implied by the feedback arrows, shown by broken lines [ 1 I ] (see Appendix, page 376).

The thrust of the assumption upon which this model is based is that variations in political systems do not explain much of the variance in outputs once socioeconomic variables are controlled. The model con- structed for this study estalishes the direction of linkages as emanat- ing from economic development to policy outcome and to the political system characteristics, and from the political system to the policy out- comes. By conceptualizing the relationship in this manner, it would be possible to assess the independent effect of the political system charac- teristics on the policy outcomes by controlling for the socioeconomic conditions, and vice versa.

In postulating causal directions it is by no means suggested that this is the only type of model possible for policy analysis. There are no

Page 5: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

365

algorithms for the purpose of differentiating the independent from dependent variables. The characterization of dependent variables is purely an identification of those aspects of social phenomena which an investigator chooses for analysis; it is not in any natural way “given” in social reality. The structure of the model produces the differentiation, although in social research there is some doubt as to which variables can be considered independent. In mathematics, algebraic operations make it possible to transpose the independent and dependent variables, but whether such transposition can be substantiated in social reality is diffi- cult to establish. The direction of causality implied in the model for this study is based upon historical data where both economic develop- ment and political system structures have been antecedent to the social policy.

The clusters of the independent and dependent variables included in the analytical scheme for this study are identified in Fig. 2. As a mea- sure of the level of economic development the principle socioeconomic input variables included are industrialization, wealth, literacy, modernity and urbanization. The characteristics of the political system are repre- sented by political participation, political competition, political stability, party representation and party system as the inputs (see Appendix, page 376).

Policy outcomes categorized into five areas of governmental action are seen as substantially being influenced by socioeconomic conditions and structures of political systems. The notion of pubIic policy, ranging in degrees of complexity, has been used in varying ways to denote actions taken by governments to accomplish some specified ends [ 121. The broadest possible usage of the term has been adopted for this anal- ysis, defining it to include the effects governmental actions have upon the population they are designed to serve [ 131. But these provisions are confined to those policies enacted for the purpose of distributing bene- fits to citizens and serving their social well-being [ 141. Operationally, the policies are measurable in terms of programs developed through the expenditure of resources of the society. Through the availability of quantitative data, as expressed in the budgetary allocations, it is pos- sible to analyze social policy systematically by the application of statis- tical operations.

Methodology and Data The dependent variable in this study is the social policy, which is

quantitatively measured by the expenditure of dollar amounts for each of the five areas of national policy in every nation. The dependent

Page 6: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

3 66

variable was operationalized in three different ways: first, the percent- age of national expenditure devoted to each of the social policy areas was calculated; and, second, the per capita expenditure in dollar amounts for each of the five areas of national policy was computed. The third set of measures included the indices for the quality of life such as the number of physicians in population, average life expectancy, number of hospital beds, infant mortality and death rates.

The two sets of independent variables include the indices for measur- ing the level of economic development, and indicators for the struc- tural characteristics of national political systems [ 15 1 . The five socio- economic variables ~ industrialization, wealth, literacy, modernity and urbanization; and the five political variables - political participation, political competition, political stability, party representation and party system, were correlated and regressed with each of the measures of the dependent variable [ 16 1 .

Analysis and Findings

Each of the 15 measures of social policy have been examined by relating them to the socioeconomic and political variables through the use of simple and partial correlation analysis. The regression model has been used to identify the best predictor variable, its coefficient of deter- mination and correlation coefficient.

Political variables, as shown in Table I, are not very strongly related to policy indicators. If a simple correlation coefficient of .SO is taken as a meaningful cut-off point, it then becomes obvious that political con- ditions are not influential determinants of social policy, since coef- ficients more than the arbitrarily selected criteria occur only 4 times, out of a possible 75. An even more dramatic change occurs when polit- ical variables are independently related to policy measures by holding socioeconomic conditions constant, where the correlation coefficients drop drastically and in several cases the relationships between the input and output variables virtually disappear. On the whole, political forces alone cannot be said to be instrumental in shaping social policies of nation-states. However, among the political variables that seem to be important in affecting social policy outputs to some degree, are political competition and political stability. It is likely that nation states enjoy- ing long periods of stability are able to enact the necessary infra-structure and implement policy decisions for social welfare measures. Such a development obviously would not take place in those nation-states where interruptions are created through extra-constitutional changes of

Page 7: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

3 67

government and political leadership. Nation-states in which there is political competition are also much more likely to enact measures for welfare of the citizens than the ones which are not. The pattcrn of correlation coefficients also suggests that some political conditions are conducive to certain policy outputs; for example, political competition and political stability seem to be much more instrumental in shaping educational and social security policy outputs than any of the others. Political conditions cannot be said to exert no influence on social policy outputs, but it can be said that the role of the political variables is not paramount, although they seem to be more important in some policy areas than in others.

A completely different picture is presented when the relationship between socioeconomic variables and policy indicators is examined. The preponderance of high correlation coefficients is striking, and this persists even when the socioeconomic variables are independently rela- ted to policy output measures by holding political variables constant. The level of the socioeconomic development of nation-states is appar- ently crucial to the structure of social policy output. It could be sugges- ted that the higher the development the more comprehensive the social policies of n a tion-s tate s.

Not all the socioeconomic variables are uniformly important for social policy output. Urbanization, on the whole, seems to be least correlated with policy variables, and per capita GNP seems to be most highly correlated. It could readily be concluded that the higher the level of per capita GNP the higher the rate of energy consumption, and the higher the level of literacy, the greater the probability that a nation would have a comprehensive system of social welfare measures. This would be true regardless of the political structure of the nation-state. If social policy outputs are accepted as programs for social betterment then the level of the socioeconomic development of the nation-state seems to be crucial for bringing about the reallocation of goods and opportunities. Although this takes place within a political framework, political forces appear to be much less important than the socio- econoinic conditions. To a slight extent, however, the competitiveness of the political system and its stability may enhance the distribution of opportunities.

A technique even more powerful than partial correlational analysis for exploring the hypothesized relationship between economic develop- ment and social policy is the multiple stepwise regression analysis which is employed in this investigation for selecting the best predictor variable. The ten independent variables were included in a single equation and regressed with each of the measures of social policy. This procedure

Page 8: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

TABL

E I

The

Rel

atio

nshi

p B

etw

ecn

Polit

ical

Sys

tem

Cha

ract

eris

tics

and

Polic

y O

utpu

t Mea

sure

s ~

~ ~

~~

~

Polit

ical

Po

litic

al

Polit

ical

Pa

rty

Part

y Sy

stem

Pa

rtic

ipat

ion

Com

peti

tion

St

abili

ty

Rep

rese

ntat

ion

Perc

enta

ge e

duca

tiona

l ex

pend

itur

e

Per c

apita

edu

catio

nal

expe

nditu

re

Perc

enta

ge s

ocia

l sec

urity

ex

pend

itur

e

Per

capi

ta s

ocia

l sec

urity

ex

pend

itur

e

Perc

enta

ge s

ocia

l se

rvic

es e

xpen

ditu

re

Per

capi

ta s

ocia

l se

rvic

es e

xpen

ditu

re

R*

PR *

* R

PR

P

PR

R

PR

P PR

0.35

0.

27

0.13

-0

.08

0.27

0.

07

0.06

-0

.12

0.06

-0

.12

0.27

0.

01

0.44

-0

.01

0.55

0.

13

0.33

0.

00

0.23

0.

04

0.24

0.

05

0.52

0.

28

0.43

0.

08

0.39

0.

20

0.28

0.

06

0.38

0.

08

0.23

0.

04

0.25

0.

03

0.50

0.

11

0.51

0.

07

0.19

-0

.03

0.37

0.

00

0.36

0.

02

0.31

0.

03

0.17

-0

.13

0.24

-0

.00

0.40

-0

.11

0.48

0.

00

0.31

-0

.06

0.14

-0

.23

Page 9: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

3 69

rn

2

2 0..

rn

8

8 m

0

0 9

QI

8

r-

8

W

0 p!

QI

0 9

4

2

W

0 3.

W x

m

8 I

r- 3

0

d

2

2 Lo

3

8

8

I

W

W

2 I

m d

0

N

2

2

I

m

m 0 9

I

3 x

c1 m W 03 9 0 I

0 2 9 0

9

3 W 0.. m 3

0 2 8 2

W z m 0 9

i i

0

r-

8 I

Lo 3

0 0

2

0.. 3

d I

d

2

N

2

2

I

0

2 0

I

a, m 0

N

2

x I

0

c1 3

0

cz

2

v)

9 .3

.I

2

N

8

2

I

0

I

c1

8

2

I

m

I

d

I 8

2 Lo

I

0 9 0

I

r-

2 I

0

I 8 ;o -4

0 I

x 3 z E

d * d

I

m

8

2

I

rn

1

d

I 8

2 d

1

3

2 I

0

0 I

3

m 2

2

I

m

I

00 z I

OI ri

0

9 0 a Q -0

V e

0

8

2

I

00

1

W

I 8

W 3

0 I

N

I 8

2 c1

I

m

8

2

I

QI

rn

0 I

4

r-

2 I

2 P i

Y .i

P z

Page 10: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

w

4

0

TAB

LE I1

The

Rel

atio

nshi

p B

etw

een

Soci

oeco

nom

ic In

dica

tors

and

Pol

icy

Out

put M

easu

res

-

Perc

enta

ge e

duca

tiona

l ex

pend

iture

Per c

apita

edu

catio

nal

expe

nditu

re

Perc

enta

ge s

ocia

l se

curi

ty e

xpen

ditu

re

Per

capi

ta s

ocia

l se

curi

ty e

xpen

ditu

re

Perc

enta

ge 's

ocia

l se

rvic

e ex

pend

iture

Per

capi

ta so

cial

se

rvic

e ex

pend

iture

Indu

stri

aliz

atio

n

R*

PR *

* __

0.40

0.

27

0.88

0.

84

0.56

0.

41

0.64

0.

50

0.45

0.

34

0.73

0.

63

Wea

lth

__

__

.

~

Lite

racy

M

oder

nity

U

rban

izat

ion

R

0.4

1

0.93

0.68

0.80

0.50

0.82

PR

R

0.28

0.

39

0.89

0.

63

0.5 3

0.

62

0.69

0.

60

0.41

0.

59

0.73

0.

63

PR

R

0.25

0.

37

0.46

0.

71

0.44

0.

57

0.40

0.

60

0.47

0.

49

0.50

0.

62

~.

PR

R

0.22

0.

24

0.62

0.

4 1

0.40

0.

34

0.4

3 0.

38

0.34

0.

28

0.50

0.

39

PR 0.1

1

0.26

0.16

0.21

0.13

0.37

Page 11: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

37 1

cn 1 0

I

3

2 I

r- 3

0 I

Lo

$

W

2

2

I

cn

I

Lo

2

2

I

N

I

\o 3

0 I

W

0 ?

I

x +- 3 z Ir 5 c c(

cr)

2

x a I

N

2

z I

d

0

0 p: I

0 p: 0

I

N d; 0

I

v) x I

01

2

x I

0

I

s 4

2 m -a

U

10 - 0' I

r-

2 I

0

I 2

x d

I

N

2

x I

W

I

00 - 0

I

rn

I 2

r- d

0 I

cn

I 2

m

B

F

P 3

c ...

z

Page 12: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

37 2

TABLE 111 The Independent Variables Which Account f o r M o s t Variance in Policy Indicators and Their R and R2 -~ - ~ - ~ _ _ _ ~

Dependent Variables Independent Variables R R2 ~ _____ _ _ ~ _ _ _ ~ . ~ _ _ _ _ ~ ~ ~

Percentage expenses for education Wealth 0.51 0.21

Percentage expenses for social security Wealth 0.74 0.55 Per capita expenses for education Wealth 0.94 0.88

Per capita expenses for social security Wealth 0.84 0.70 Perccn tage expenses for social services Literacy 0.63 0.40 per capita expenses for social services Wealth 0.82 0.68 Percentage expenses for transportation Wealth 0.27 0.07 Per capita expenses for transportation Wealth 0.70 0.49 Percentage expenses for public health Political participation 0.19 0.03 Per capita expenses for public health Wealth 0.74 0.55 Physicians Literacy 0.87 0.76 Life expectancy Literacy 0.85 0.73 Infant mortality Literacy 0.52 0.27

Literacy 0.74 0.55 Crude death H o spi t a1 Literacy 0.48 0.23

~ ~ _ _ ~~ - ~~ ~~ -

made it possible to select the single variable which accounted for most of the variations in each of the policy indices, and to determine whether this was a socioeconomic or a political variable. Table 111 shows the type of the independent variable that contributed most to an explanation of variance in the dependent variable, its correlation coeffi- cient and coefficient of determination.

The best predictor variables are found to be the socioeconomic ones; only in one case did a political variable show up as an explanatory vari- able. Even in this instance it has the lowest coefficient of all the varia- ables that entered the equation first. On the average, a socioeconomic variable accounts for 51 percent of the variation in a single dependent variable whereas the political variable accounts for only 3 percent of the variation.

Conclusion

The findings presented in these pages overwhelmingly sustain the hypothesis that the level of economic development is the prime deter- minant of social policy. Political factors, at least statistically, are found to be less than crucial for social policy outputs. It is likely that political conditions are instrumental for shaping policies in many other areas, but in the realm of policies dealing with collective welfare, other con- siderations seem to be more important.

Page 13: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

373

Furthermore, variations in socioeconomic conditions seem to account best for variations in social policies among nations. Variations in political characteristics, while considerable, did not seem to be salient in explaining variance in social policies. Moreover, socioeconomic vari- ables are found to be the best predictors of variations in social policies. Socioeconomic variables, on the other hand, more often showed con- sistently high correlations with policy indicators. While not exhibiting such strong relationships, two political variables also portrayed a rela- tively higher association with policy output indices. These variables, political stability and political competition, evidently are important for social policy, and are also highly correlated with economic development measures. On this basis alone it could be accepted, that other things being equal, a highly developed nation which is stable and has a com- petitive political system would also have a high level of expenditure for social policy. In other words, the level of social policy could be predicted, once the socioeconomic and political conditions of a nation are known.

Wealth and literacy seem to be the two most important socio- economic variables in explaining social policy variance among nations. Perhaps it can be concluded that ultimately wealth helps in the devel- opment of welfare for citizens regardless of the political complexion of a nation.

These findings are in conformity with conclusions reached by several scholars working in the field of American state politics. This is all the more noteworthy since the data for the present study, derived from a cross-cultural setting, sustains the hypothesis that socioeconomic devel- opment is instrumental in structuring the social policy outputs of nation-states. There is a recurrent pattern of strong relationships be- tween socioeconomic conditions and policies designed to enhance citi- zen welfare. Variations among nation-states in these policy outputs seem to be accounted for by difference in socioeconomic conditions. There are two socioeconomic variables that are particularly impressive in showing exceedingly strong relationships with policy measures, and also in accounting for policy differentials among nation-states. These are wealth and literacy. It seems plausible to suggest that wealth and literacy are central in explaining why some nation-states have forged ahead in providing for opportunities and redistribution of benefits through education, social security, social services, transportation and public health, while others have not.

Though variations in social policy outputs are closely related to differences in the level of socioeconomic development among nation- states, it can also be suggestively concluded that socioeconomic devel-

Page 14: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

374

opment shapes both the political systems and policy outputs, and most of the association that occurs between political variables and policy measures can be attributed to the influence of socioeconomic factors.

1 Richard E. Dawson and James A. Robinson, “Inter-Party Competition, Eco- nomic Variables, and Welfare Policies in the American States,” Journal of Politics 25 (May, 1963), 265-289; Alan K. Campbell and Seymour Sacks, Metropolitan America: Fiscal Patterns and Governmental Systems (New York: The Free Press, 1967); Thomas R. Dye, Politics, Economics and Welfare (Chicago: Rand Mc- Nally, 1966); and Ira Sharkansky, The Politics of Taxing and Spending (New York: Bobbs-Merrill Press, 1969). These volumes have generally explored public policy as outputs shaped by socioeconomic variables.

2 V.O. Key, Jr., Southern Politics (New York: Alfred E. Knopf, 1949) and Duane Lockard, New England State Politics (Princeton: Princeton University Press, 1959) emphasize the primacy of political conditions as the determinants of public policy.

3 The systems conceptualization is based upon David Easton’s A Systems Analysis of Political Life (New York: John Wiley, 1965).

4 Richard Dawson and James Robinson, op. cit. 5 Thomas R. Dye, up. cit. 6 Philip Cutright, “Political Structure, Economic Development and National

Social Security Programs,” American Journal of Sociology 7 0 (March, 1 9 6 3 , 537-548.

7 Alexander Groth, Comparative Politics: A Distributive Approach (New York: MacMillan, 1971).

8 May Brodbeck, “Models, Meaning and Theories,” p. 583 in May Brodbeck, (ed.), Readings in the Philosophy of Social Sciences (New York: MacMillan, 1968).

9 This idea is taken basically from the works of David Easlon. 10 In a related context this idea is used by Henry A. Kissinger, “Domestic Structure

and Foreign Policy,” p. 262 in James N . Rosenau, (ed.), International Politics and Foreign Policy, rev. ed., (New York: The Free Press, 1968).

11 Since statistical routines are applied t o an analysis of data, the model may be mathematically stated as follows:

D = e l P = K I D + e2 S = K2D + K3P + e3

where

D is the level of socioeconomic development P is the political characteristics S is the expenditure on social welfare programs e represents error or variable left out K represents constant

Adapted from Charles F. Cnudde and Donald J. lClcCrone, “Party Competition

Page 15: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

37 5

and Welfare Policies in the American States,” American Political Science Review 6 3 (September, 1969), 855-859.

12 Theodore J . Lowi, ”American Business, Public Policy, Case Studies and Political Theory,” World Politics 16 (July, 1964), p. 689, conceptualizes policies in terms of “their impact or expected impact on the society.” Harold Lasswell and Abraham Kaplan, Power and Society (New Haven: Yale University Press, 1959), p. 71, argue that “policy is a projected program of goals, values and practices.” For Wilfrid Harrison, “Public Policy,” in Julius Could and William L. Kolb, (eds.), A Dictionary of Social Sciences (New York: The Free Press, 1964), p. 509, the term policy “refers to a course of action or intended course of action conceived as deliberately adopted, after a review of possible alternatives, and pursued, or intended to be pursued.” Policy is viewed by James A. Robinson, Congress and Foreign Policy Making (Homewood: Dorsey Press, 1962), p. 3 , as “goals of any social system, the means chosen t o effectuate those goals, and the consequences of the means, i.e., the actual distribution of values.”

13 Austin Ranney, “The Study of Policy Content: A Framework for Choice,” in Austin Ranney (ed.), Politicul Science and Public Policy (Chicago: Markham, 1968), p. 7. See also Robert Lineberry and Ira Sharkansky, Urban Politics and Public Policy, (New York: Harper and Row, 197 I ) , Chapter 6.

14 Theodore J . Lowi, “Distribution, Regulation, Redistribution: The Functions of Government,” in Randall B. Ripley, (ed.), Public Policies and 7’heir Politics, (New York: W.W. Norton, 1966), pp. 27-40.

15 The socioeconomic variables are measured in the following way: industrializa- tion by energy consumption per capita; wealth by per capita GNP; literacy is represented by the percentage of the population over 18 with the ability to read and write; modernity by the percentage of the labor force engaged in non- agricultural economic pursuits; and urbanization by the percentage of the popu- lation defined as living in urban areas by each of the nation-states. For political variables: political participation is represented by the percentage of the eligible voters who voted in the last national election; political competition is based on an index constructed by Arthur Banks (see footnote below) for each of the political systems included in this study; political stability is represented by the number of years since the adoption of the last constitution; party representation is the pattern of representation in the national legislature, with a zero value for no party t o a value of 4 for a majority-minority relationship; and party system is the number of political parties extant in a nation-staie that have had national representation within the last ten years.

16 Since the interest in this study has been on cross-national analysis, data from 125 nations that were sovereign in 1968 and had been so for a t least three years were selected for examination. Quite a variety of data sources were consulted, with the United Nations compilations providing the bulk of the figures used. A considerable proportion of the expenditure data was obtained from the U.N. Yearbook of National Statistics, and the U.N. Stutisticul Yearbook, the rest from such diverse sources as The Ezrropa Yearbook, publications of several inter- national organizations, and statistical offices of various nations. Most of the socioecononiic data was derived from the U.N. Demographic Yeur- book, Hudson and Taylor’s World Handbook of Political and Social Indicators, and the yearbooks of UNESCO, ILO, F A 0 and WHO. For data on political variables, the sources consulted were Arthur Bank’s A Cross

Page 16: POLITICS, DEVELOPMENT AND SOCIAL POLICY: A CROSS-NATIONAL ANALYSIS

376

lndustr ia l izat ion Wealth L i t e r a c y

Pnliticul Survey and Cross Polity Time Series Datu, Politicul Hurzdhook of the World, Review of Electioiis and Hudson and Taylor’s World Haridhook ofpolitical and Social Indicutors. Virtually all the da ta used in this study was for the year 1968.

Appendix

~-

Fig. 1. A model for analyzing policy outputs.

SOCIOECONOMIC VARIABLES

Fig. 2. The variables included in the policy analysis model.