Francis Castles goes for the high stakes in political sociology. Comprising 162 pages his volume '[e]ndeavours to explain the political dominance of the Scandinavian Social Democratic parties and to demonstrate the consequences of that dominance.' (p. x). Such undertakings have, of course, all the possibilities of falling short of the target. No two readers of the book will agree in full with the way Castles emphasizes and plays down the explanatory factors. And (too) many readers will suggest (too) many additional variables in assisting Castles in his painstaking task. In my review I shall, therefore, refrain from acting like an insatiable computer processing a never-ending stepwise multiple regression analysis without a cutoff criterion.

1. Setting the Outline

Castles' introductory chapter '[a]ttempts to describe the nature and magnitude of the basic contrasts in the electoral situations of the European democratic socialist parties . . „ and sketch the various stages of development through which each of the Scandinavian labour movements has acquired and maintained its political ascendancy.' (p. 4). Contrary to his intention, Castles offers interesting explanations to phenomena he merely claims to be describing.

'Three aspects of the development of the Scandinavian labour movements have been particularly influential in conditioning their strength, unity and integration. These were the relative absence of impediments to working-class industrial and political organisation, the timing and social context of organisational growth, and the nature of the strategic choices faced by the various movements at different times.' (p. 13)

The relationship between bourgeois repression and arithmetical strength of organized labor (Social Democrats and Communists) is not clear-cut. The liberal U.S. Constitution is said to have erased the raison d'etre of socialism in America. Bourgeois repression and labor unity, on the other hand, correlate quite strongly. In Finland, France, Italy, and Weimar Germany the hard line of the bourgeoisie boomeranged and contributed to strong Communist parties. The Norwegian Labor Party represents the exception to that rule. It joined the Comintern in spite of Norway's liberal Constitution of 1814.

Castles handles the recurrent topic of why Labor in Norway turned left in 1918 with admirable lack of respect for the somewhat sacrosanct 'Bull-Galenson hypothesis'. This hypothesis states that Labor's radicalization stemmed from the late but sudden and rapid industrialization of Norway. To this Castles rightly comments: '[t]hese views [cannot] be unequivocally accepted, if only because economic historians provide evidence which suggests that, to the extent that there was a difference in Norway and Sweden's growth rates in the first two decades of this century, it was Sweden's that was the most rapid.' (pp. 19-20). Castles prefers to bring strategic necessities to the fore when analyzing the different courses taken by Scandinavian Social Democracy: 'The absence of the basic prerequisities of democratic participation in Denmark and Sweden forced the Social Democrats into a prolonged collaboration with the more progressive sections of the bourgeoisie.' (p. 21).

In a similar vein, social and political circumstances called upon Social Democracy to act as a strategic pivot in the aftermath of the 'Big Crash' in 1929. 'Crisis Agreements' were reached between the Social Democratic and the Agrarian parties in Denmark, Sweden, and Norway, in that order. Here Castles, trying to find programatical support for the Social Democratic posture, is dangerously close to committing an ex post rationalization - the agreements did imply an ideological about-face as far as the Social Democrats were concerned. Castles then subscribes to the common notion that the realignment of the 1930s was a watershed in the history of Scandinavian Social Democracy. From then on the party was '[t|o institutionalise mechanisms for collaboration between organised social groupings.' (p. 30).

Though it does not alter the overall impression of Castles' historical overview, it is annoying to note that the author has accepted the myth that Denmark has been a nightwatch state (p. 33). Available statistics bear no support for that. Furthermore, Castles falsely credits the Swedish farmers with having contributed to '[t]he democratisation of the political systems in the late nineteenth and early twentieth centuries.' (p. 114). The truth was rather the other way around.

2. The (Im)potency of Politics Revisited

In chapter 2 Castles sets out '[t]o show not only that the level of welfare state provision in Scandinavia is exceptionally high, but also that the most satisfactoiy explanation lies in the political ascendancy of Social Democratic parties in those countries.' (p. 57, emphasis in original).

'Welfare', one of the most (ab)used catchwords of our time, carries connotations like 'well-being' and 'equality'. The former concept has but little to do with welfare. Apart from being laden with subjective arbitrariness, a nation's well-being' tells us nothing about the distribution of purportedly standard-raising goods and services among individuals. Equality, on the other hand, is welfare. If we could prove by empirical observations that a child of a working-class family and a child born into bourgeois affluence had the very same prospects to reach whatever goals of his or hers, the level of welfare would be absolute (and the topic an anomaly in the first place). With aggregate data on the lurid side, there are infinitesimal possibilities that each individual has a fair chance in life expectancy, especially in terms of quality.

However, it is financially unthinkable to conduct large scale panel-studies on the social history of entire nations. The large surveys into the 'level-of-living' of Scandinavians were all of the aggregate variety and not based on panels. In effect, scholars have been forced to rely on second-best data (i.e., aggregate data) when studying welfare among analytical units such as nations, regions, communities, etc. In this process of data collection there has been a tendency to blur the initial implication of 'welfare'. The non-existence of sheer misery, or a minimum government guarantee of decent living conditions, has become a substitute for equality. Though publicly provided welfare programs tell us little about egalitarianism as far as each individual's outlooks are concerned, they are an indicator of how many, and how great, the hazards a citizen will be subjected to in a lifespan.

Castles presents four variables measuring the absence of sheer misery among OECD countries: (1) government revenue as percentage of GDP, (2) public spending on education as percentage of GNP, (3) infant mortality, and (4) GDP per capita. His data pertain to 1973 or 1974. With the first three variables cast as a combined 'Index of Pure Welfare', Sweden and the Netherlands score a tied no. 1 position, followed by Norway and Denmark in that order. Finland comes in as no. 5, and Canada no. 6. At the bottom of the ranking list we find countries like the U.S., Ireland, and Italy (see Table 1). With regard to the clustering of nations on the respective extremes of the list, the outcome probably surprises nobody.

However, the choice of factors to go into the index is decisive for how the
countries fall out on the relative welfare scale. In fact, different sets of variables
produce different overall welfare rankings of the nations.

West-Germany and Austria highlight the difficulty inherent in cross-national welfare comparisons. On Castles' ranking list West-Germany turns out as no. 12 - on mine as no. 2; Austria scores a meager 13th position on Castles' list- on mine that country comes in as no. 3. The Scandinavian countries' performance is not overly impressive according to my alternative index. Sweden and Denmark occupy positions no. 5 and 6 respectively. Norway, on the other hand, clearly turns out as no. 1, and defends the lead when Castles' data are incorporated into my welfare indicators (see Table 1). Thus, Castles' statement that '[t]here does seem to be clear evidence . . . that the Scandinavian Social Democratic claim that 'we're far ahead compared to other countries' is justified' is a somewhat foregone conclusion (p. 73).

Now, I would be the last to claim that my welfare indicators are any better - or worse, for that matter- than those suggested by Castles. Data on unemployment might, if only they were reliable, be the measure of welfare. It has been documented time and again that being out of work has devastating effects upon individuals. Castles refrains from including social security into his index on the grounds that such spendings are '[mjainly of a social insurance character.' (p. 72). But that is precisely what welfare is all about when tackled at the aggregate level. Social security and infant mortality are basically the same side of the coin. The difference is chronological: social security is supposed to prevent adult mortality. Besides, infant mortality has become somewhat of a welfare showcase. It is relatively easy to keep comparative records on the number of children who are unfortunate enough not to reach one year of age. Some politicians have, therefore, allocated large shares of their limited resources on reducing infant mortality partly

for the reason not to appear on the dark side in international statistical yearbooks. The inclusion of the strength of the coercive apparatus into my welfare index is warranted simply because military materiel and personnel detract from national resources; and the number of policemen is a hint about 'non-existency of sheer misery' in a society. After all, delinquency is strongly related to poor social conditions.

One cannot avoid feeling that Castles has been too biased when finally deciding
upon which welfare indicators to be included in his book.

Castles defines the independent variable (political structure) in a very sweeping way: Scandinavia has been ruled by Social Democracy, and the rest of the West by bourgeois forces. The fact is, however, that in none of the Scandinavian countries have the Social Democrats had uninterrupted periods of calling the shots. Even in Sweden the party had to cooperate in a coalition with a bourgeois party in order to remain in power (the 1951-57 Social Democratic-Agrarian cabinet). Both Denmark and Norway have experienced bourgeois governments during the postwar years, i.e., in the late '60s and early '70s when, ironically enough, '[t]he percentage of Gross National Product devoted by each of the Scandinavian countries to social security increased by more than a half.' (p. 59). But Castles' comparative inquiry into the finer details of welfare state provision in Scandinavia is not systematically cast with an eye to which parties were in power. Such a juxtaposition of data sets, however, would not, as should be obvious by now, produce evidence in support of Social Democratic dominance of welfare reform legislation. In this light it is surprising that Castles fails to mention the relationship between welfare and politics at the local level. Although the correlations between party strength and welfare spendings among local governments in Sweden have been on the low side, there are intimations which support Castles' thesis.

3. Explaining the (Im)potency of Politics

Castles wraps up his volume with an '[a]ttempt to explain the singularity of
contemporary politics in Scandinavia in terms of a syndrome of structural and
historical circumstances which only they have in common.' (p. 105).

Undoubtedly, part of that syndrome is the straightforward (read: left-right) Scandinavian conflict structure: 'There are no substantial minorities based on religion, language or ethnic origin in Scandinavia, which in many European countries have formed the basis of political parties with a cross-class appeal, and have offered an alternative focus of loyalty to sections of the working-class vote.' (p. 106). The Low Countries are good cases in point as to how social and cultural fragmentation have impeded the strength of organized labor, both the political and industrial branches.

However, there is more than extrapolation of working-class electoral behavior to the success of Scandinavian Social Democracy. And that something is still more than '[t]he fact that they [the Social Democratic parties] can make a cross-class appeal as a party of the whole nation.' (p. 112). By and large, non-working class voting for today's catch-all Social Democracy may be explained with reference to socio-psychological variables, i.e., the class and partisan milieu in which the voters were brought up.

What determined the ascendancy of Social Democracy in Scandinavia, and left

its fellow comrades in the rest of Europe helplessly trailing, must be traced way back to the 1930'₅. Ecological data, and later sample survey data, have testified to the extreme stability of the Scandinavian party systems ever since the late thirties. In his introductory chapter Castles also repeatedly stresses this very decade when pointing to the differences among the Social Democratic parties' electoral records in contemporary Europe (pp. 7-8, 12-13,26-27). But in the closing sections of his book Castles more or less forgets about the broader comparative perspective. Instead he focuses on within-Scandinavian similarities, in contradistinction to the British case. An analysis of why and how the Scandinavian Social Democrats - as opposed to the sister parties in Continental Europe - managed to escape from their ideological straightjackets in the 1930s and attract non-working class strata and a bourgeois party would shed light on fundamental differences in the social and political systems of Europe. Are, as Castles touches upon, '[compromise, stability and harmony . . . characterising features of the Scandinavian societies' (pp. 118-9)? And, if so, did those attributes have anything to do with the 'Crisis Agreements' and subsequent political horsetrading?

As to the 'dividing of the cake', Castles' impression of Scandinavia may be summarized by the above mentioned concepts. 'The politics of virtuous circles', as Castles puts it, have ensured stable progress to the benefit of the large majority of the Scandinavian people. 'The reciprocal interrelationship between the maintenance of Social Democratic political ascendancy, labour movement solidarity in pursuit of economic growth, and the achievement of the twin goals of full employment and enhanced welfare state provision is but one aspect of the politics of virtuous circles. It is, however, the aspect that should be stressed most strongly In comparative perspective, since the mechanism involved is that which most clearly distinguishes Scandinavian political life in the last four decades from that elsewhere, and which has functioned as the precondition of the more extensive social and economic co-operation that has characterised Scandinavian social organisation in the same period.' (p. 128). As a description of Scandinavian politics this is not too far off the mark. With the possible exception of Denmark, industrial conflicts in Scandinavia are something of the past; and the trade unions have, in the 'national interest', often been willing to settle for less than their 'due' when signing contracts. Castles interprets this labor market tranquility as the definite proof of the confidence that the working-class puts in its party! It does not make much sense to launch a strike because, in the end, the Social Democrats will see to the lot of the working-class. But the fact stands out that the government, i.e., the Social Democratic Party, has outfoxed the unions on several occasions. In Sweden, for instance, the Social Democrats introduced VAT despite heavy protests from the unions. In the mid-seventies, when the unions agreed to wage increases amounting to nickels and dimes at a time when business dividends skyrocketed, the Social Democratic government set up a new investment fund in which the companies could place their surplus in interest-earning accounts that were exempt from taxation. No union member ever came close to that money. Such economic maneuvers would hardly succeed if it were not for the close ties between the unions and the party at the elite level. Corporatism in Scandinavia has gone so far that the unions will hurt their own vested interests by going on strike. Moreover, if there indeed exists a mutual trust of confidence between the workingclass rank-and-file and the elite, why are the unions unwilling to let their members

vote on the terms of the contracts? Not with one word does Castles mention that
the overlapping of high office in the party and the unions implies conflicts of
interest.

4. Conclusion

At the outset, Castles faced two major avenues open when tackling his topic: (1) a within-Scandinavian or (2) a general Western comparison of welfare state provision. The former approach would have required a longitudinal design in order to obtain some variance in the independent variable (cabinet composition), and possibly also an analysis of the relationship between policy and outcome at an intermediate governmental level. The general Western comparison, which Castles settled for, derailed en route.

The book winds up as a discussion of Scandinavian similarities and differences, with occasional glances at Europe in general and Britain in particular. A comprehensive set of welfare indicators among Western nations, both in quantity and quality, would have permitted Castles to make more convincing conclusions about the Scandinavian countries' welfare output. After all, the public sector in Scandinavia is considerably larger than that of other Western countries (p. 68), and it would be far-fetched to assume that the Danish, Norwegian, and Swedish governments pour the taxpayers' money down the drain (that includes feeding the bureaucracy)! Having a rather weak battery of welfare data, Castles was hard put to arrive at authoritative explanations as to why Social Democracy has fared so much better in Scandinavia compared to other nations. Parts of the section 'The politics of virtuous circles' were a heartbeat from mere (democratic) socialist polemics of dubious circles.