Non-voting in Spain : some quantitative and attitudinal aspects

Non-voting in Spain: some quantitative and attitudinal aspects José Ramón MONTERO Universidad Complutense de Madrid Working Paper n.22 Barcelona 1990 After five legislative elections have been held, the main features of the electoral behaviour of Spanish voters can be approached with some certainty. The electoral non-participation will be analysed in this working paper which aims to review some of the hypotheses and generalizations that were made on non-voting when just a few elections had been held. We shall base our research on two specific aspects. On one side, the quantitative aspects that will allow us to compare the Spanish non-voting with that of other European countries with regard to some basic dimensions, as well as that of the regions (Autonomous Communities) among themselves. On the other, the attitudinal aspects that meet in certain cultural contents of non-voting and in the non-voters' mobility. Levels, fluctuations and trends of non-voting Generally speaking, we can qualify the Spanish non-voting levels as being high. The variations between different elections are also considerable as well as fluctuations in time between elections of the same kind. Percentages of non-voting in the 23 elections held between 1977 and 1989 have not gone, except in a few occasions, below 30%, (see Table 1). Differences related to the kind of election ratify the hierarchical order that voters construct in relation to the importance given to each of the electoral contest: the national are "first-order" elections, followed by the regional, local and, with a much lesser prominence, the European elections (1). The simultaneous celebration of different kinds of elections has nevertheless occasionally put on the same level non-voting rates distinguishable a priori. This was the case, for example, in 1986 when the national elections coincided with the regional in Andalucía, or in 1983 and 1987 when the local (and European in 1987) elections coincided with the regional in thirteen regions. Those considered of higher relative importance had a towing effect over the rest, obtaining similar non-voting rates. Among the "second-order" elections, there seems to exist a trend towards a lower non-voting in the local elections, while any conclusion about the European and regional elections would be premature because of the contradictory oscillations, the short number of cases or/and the exceptional circumstances of simultaneous election days. In any case we must say that the 45% of non-voting in the elections to the European Parliament in 1989 is not much different from the rates obtained in the other eleven European countries; but the rates in the Spanish local, and even more regional, elections are usually higher than those of countries with comparable elections (2). Table1 Spanish non-voting is one of the highest in Europe. In fact, Spain is the second in a list of the eighteen European countries ordered by their non-voting averages since 1970 (Table 2) (3). If we leave aside the Swiss case, with really exceptional nonvoting levels, the countries with a higher non-voting average are Spain, United Kingdom, Ireland and France, followed in a short distance by Finland, Greece, Portugal and Norway. It is remarkable, that the non-voting higher rates are shared by countries that belong to such unlike geographic areas as Southern Europe, the British Islands or the Scandinavian Peninsula: this is indicative enough of the complexity of the non-voting phenomenon. Another remarkable fact is that these same countries show in general a trend towards the non-voting increase and moreover they do so in a greater extent than the more participative countries. In some cases (France, Finland, Portugal) the celebration of presidential elections may have reduced the importance of the parliamentary ones, which could suffer the demobilization effects of the "second-order" elections. In others (such as Ireland) the confluence of historical, cultural and structural factors has obstructed the electoral participation since the introduction of universal suffrage. The British case has been qualified as exceptional because the increasing trend of its non-voting in the last half century coincides with the development of the political information and the electoral competition, on one side, and with the fall of party identification and electoral stability, on the other (4). Table2 Spanish non-voting was, on the contrary, high from the first moment: its average level surpasses that of the so-called Southern European democracies and that of the political systems that: have recently gone through crises and breakdowns. In Spain it has not been possible to reduce non-voting below the 20% obtained in the founding elections in 1977 when a "participative explosion", was expected, similar to the experienced by other countries with an authoritarian past (5); neither in 1982 when their condition of critical elections (and, incidentally, the making up of a new census) predicted a greater mobilization (6). Thus there is a feeling that the Spanish voting levels are producing suddenly results that in other contexts are perceived as the consequence of evolutive processes. That is why it would not be exaggerated to remark that one of the distinctive elements of the Spanish democracy lies not so much in its late arrival to the European scene but in its modernity. In other words, in developing a certain pattern of attitudes and behaviours towards which other European political systems are going but have not reached yet (7). This paradoxical modernity, a consequence of the democratic system's birth conditions, is also present in other dimensions of Spanish electoral behaviour: decrease of partisan alignment, progressive fall of party identification and increasing importance of leadership are some significative examples. Table 2 also contains some information about the non-voting fluctuations, shown by the differences' dispersion around the mean and measured by the standard deviation. The greater variations usually occur in those countries with higher non-voting rates. They have fluctuated between seven and thirteen percentage points in Portugal, France, Finland and Spain (Table 3). While in the first three of these cases fluctuations have been sporadic -maybe due to their relation with the presidential elections dynamics-, in Spain they have taken an almost cyclical form. Since the first elections in 1977, the non-voting rates have suffered constant increases or decreases of around ten percentage points, with the one exception of the last 1989 ones. To appreciate the importance of these fluctuations we must consider that voting and non-voting rates are usually characterized by a, considerable stability. The non-voting rates' differences that exist between countries seem to be compensated by the scarce variation within each of them. Just five countries among the eighteen selected for a comparative study have known variations equal or higher than 5% from the 1940s to the 1970s. The - variations' mean experienced between consecutive elections is of 2,7%, a part of which can be attributed to the census' imperfections. Consequently, a variation of 5% in the non-voting rate means a certainly important change (8). Among the little more than one hundred elections that have been held in the eighteen European countries selected, this variation equal or higher than 5% has just appeared in twenty cases; and they have affected, as shown in Table 3, eight countries, among which, as we have already said, Spain occupies an outstanding place. The fact that an important number of these variations has occurred within the last years seems to reinforce the hypothesis on the existence of an increasing electoral volatility not only related to the change of vote among the different parties but also to the very, decision of voting or non-voting in successive elections (9). And thus many of the countries selected are now experiencing positive trends in the non-voting rise, a fact that also appears in Table 2. The highest ones usually correspond to those countries with higher non-voting rates. Nevertheless, we must take into account the different rates on the basis of the positive and negative trends, as is shown in Figures 1 and 2. The non-voting trend in Switzerland cannot be put on the same level that the West Germany's, for example, in a similar way as how the British or Greek falling trends have very different departure points from the Belgian. Table3 Figure1 Figure2 In the inner Spanish context time has confirmed the trends already noticed in the first democratic elections. As Table 4 shows, the non-voting average level in Galicia continues to be exceptional. If we consider the most non-voting regions, the complexity of the phenomenon outstands once more: Galicia and the Canary Islands, Asturias, the Basque Country and Baleares, Catalonia and Andalusia share high non-voting rates despite the economic, social, cultural and, of course, political differences they show. And something similar occurs, in the other extreme, with the most voting regions, such as Rioja and Madrid, Castile-Mancha and Valencia. The geographical criterium has usually been useful to show some basic structural aspects from which we can classify those areas with a higher or lower tendency towards non-voting. In this sense several studies have coincided to indicate the higher non-voting rates since 1977 in the coast and island regions, and the lowest in the interior regions, with the exception of Cantabria and Levante. This generalization is also good for the districts and even for the main cities (10). Table4 Non-voting fluctuations at national level have reproduced in the regional and district levels. Non-voting percentages in all the regions and the majority of provinces have developed along with national rates. Table 5 shows a first sample of the scarce differences appeared in the electoral geography. The coefficients of correlation for the non-voting percentages' territorial distributions are very high in the districts and, though in a lesser measure, in the main cities. This reinforces the similarity between the oscillations suffered in the districts (11). An additional proof consists in the lack of a direct relation between higher non-voting levels and greater oscillations between successive elections: the fluctuation rate -as shown in Table 4is as high for the most nonvoting regions as for the most voting ones. Something similar happens in relation with their respective trends. They both show the particular changes that seem to be going on in some regions. For instance, Baleares, which has become since 1986 one of the most nonvoting regions; or Extremadura, which has moved from the fourth place among the most non-voting regions in 1979 to the second place among the most voting in 1989; and even Madrid, the most participative one in 1979 and 1982, and known, in relatively terms, as the Castilian meseta's "non-voting island" in 1986 and 1989 (12). Figures 3 and 4, which reflect the negative trends of Galicia and Extremadura and the positive trends of Aragón, Baleares, Navarra and Madrid, show the disparities we are talking about. Table5 Figure3 Figure4 These changes coexist with another two trends that must be at least pointed out. The first one shows an increasing reduction of the differences between the regions' non-voting rates (Table 6). The interregional homogenization evidences itself in the smaller range of the non-voting percentages in different situations. The second trend has to do with a greater intraregional homogeneity that seems to be taking place in most of the regions. The oscillation rates' evolution in the non-voting levels of the pluriprovincial regions' districts shows also a I growing similarity (Table 7). Although there are some exceptions (as in Andalusia) and some oscillations (as in the Canary Islands and Aragón, among others) the general trend in nine of the ten selected regions goes towards the sometimes considerable reduction of the intrarregional differences that were clear in 1977. Thus we could talk of a certain "nationalization" in the Spanish electoral behaviour's participative dimension, in the sense of higher homogeneity levels between different regions as well as intraregional in those with two or more districts. Table6 Table7 Voluntary and unvoluntary non-voting Apart from its quantitative aspects, non-voting evolution and magnitudes in Spain have had a significative qualitative projection among the political elites and the electoral analysts. Their hypotheses centered around the non-voting incidence on the transition and the democratic consolidation processes, and focused on the non-voting types, its causes and the nonvoters' profile. Many observers considered the non-voting increase since 1977 as a worrying disaffection index towards the new democratic regime. Others saw its extraordinary fall in 1982 as an indicator of a new legitimacy that would permit the system's definitive consolidation. In the European electoral processes, nevertheless, the causal relation between participation and legitimacy (so that a considerable fall of the first would reduce similarly the second, and vice versa) is not very valid; and its exceptions are too many as to be confirmed in the empirical theory of democracy. But this relation seemed to have a general validity for the interpretation of the elections between 1977 and 1982. The nonvoting increase was considered as equivalent to the political system's loss of legitimacy, the weakness of parties institutionalize or the popular isolation towards their leader. Non-voting was causally seen as an effect of "disenchantenient"; and non-voters were related with some supposed antidemocratic attitudes that were dangerous for the democratic consolidation. That is the reason why, after the coup d'état attempt in 1981, the higher electoral participation in 1982 was understood as a kind of plebiscite in favour of democracy: a plebiscite that finished with the consolidation crises that occurred under the Unión de Centro Democrático (UCD) governments and opened the way for the PSOE's governmental alternative (13). Since 1986, 30% of non-voters has usually been simultaneously interpreted as an assimilable rate to that of the European countries and as the homogeneous expression of the lack of political preferences of a supposed "non-voters' party". Since the first interpretation has already been pointed out, we will focus on the second. In fact, any reference to non-voting and to non-voters should be always done in plural,(14). Thus the non-voting rates' oscillations have depended on the concurrence of the structural factors that characterize districts with the occasional factors of each election. Among the latter we can mention the exceptional conditions (or conditions for change) of the 1977 and 1982 elections, and the normality (or continuity) conditions of the 1979, 1986 and 19.89 elections, to which an important part of the participation's rise or fall (and maintenance) can be respectively attributed (15). From another point of view, the nonvoting typologies consider many hypotheses. The most important ones distinguish between the voluntary and the unvoluntary non-voting. The voluntary non-voting motivations are usually grouped as active (or of protest) and passive (or of indifference), while the reasons of the unvoluntary generally come from work problems, census' defects or difficulties linked to health or age (16). We lack any empirical research that deepens in these hypotheses and offers quantitative estimates for the Spanish elections. We can nevertheless point out some interesting facts. When we ask about the non-voting motives to a representative sample of the population in a postelectoral survey, the majority of answers outstand lack of interest, indecision and indifference. But when non-voters themselves answer to this same question, the reasons of unvoluntary non-voting reach considerable rates (Table 8). It does not seem to exist an uniform guideline for the distribution of opinions about the causes of non-voting in national or regional elections, except perhaps for the general one of those who justify their lack of interest on vote with technical or force majeure reasons (17). Table8 In spite of this, the voluntary non-voting acknowledgement is also high. In the democratic systems the dominant assumption of voting as a civic duty makes non-voting a reprovable behaviour which is why many non-voters conceal their attitude in, survey questions or justify it with unvoluntary reasons. But it is likely that the agitated suffrage's history in Spain and the social conditions of its recent establishment have weakened the cultural assumption of the obligation of voting. Some two thirds of Spaniards do vote because they consider it a civic duty (Table 9). This is a high rate but it is still far from the overwhelming percentages -near to 90%- in other countries (18). And something similar happens with the satisfaction feelings linked to the vote, which are loudly low in Spain (19). Besides, this lower cultural and emotional projection of vote is reinforced by a higher acceptation of non-voting as a legitimate political expression form (Table 10). Comparatively with other Southern European countries, Spanish and Portuguese condemn non-voting less forcefully than Italians and Greeks do, and show, on. the contrary, a more marked understanding attitude towards it (20). The fact that the social groups that seem more understanding are the younger and with higher education levels seems to point out towards a future increase of the non-voting legitimacy. So, the Spanish case is also in this aspect set aside from the European countries, where the voting civic duty or/and the social pressure for voting explain high participation levels despite a slim interest in politics or a low sense of political competence (21). Table9 Table10 Among the various voluntary non-voting types it is very likely that the positive or active is certainly a minority. It consists in expressing, by non-voting, an opinion on the political system, the structure of electoral choices or the dynamics of partisan competition. The active non-voters -with a higher ideological and political interest level than the averagehave very different motivations, generally critical, with which they want to influence the political life. In 1989 a significative section of the Spanish population considered that non-voting was justified as a protest form (Table 11). Although this is the motivation chosen by almost half of the non-voters in Table 11, in postelectoral surveys the protest or critical option is adduced by a much lower rate (22). Spaniards do not generally show significative differences when they relate non-voting to a group of motivations, which include lack of, party attraction, non-voters' irresponsability or a protest expression. But non-voters themselves seem to vary a little this diagnosis by emphasizing the importance of the party factors and logically reducing their irresponsability doses. The resulting image fits without any difficulty with the nonvoting social bases: the majority of non-voters correspond to the passive type. As in other Western countries, their absence at the polls is mainly due to social marginality, political apathy, lack of interest, lack of information and indifference (23). Many studies have confirmed also in Spain higher non-voting levels among women and housewives, the youngs and the old persons, unmarried, unemployed, pensioners and students; among those with lower studies levels, lower income, who do not belong to any voluntary society, declare to be good catholics, atheist and not interested in religion; and among those who .lack any information about the political events, who consider themselves completely out of the political life or who are not interested at all in politics (24). Although with some exceptions of secondary importance, these sociodemographic profiles of non-voters are substantially the same at the regional level. Table11 As for the unvoluntary non-voting, the principal types are related to force majeure reasons, to the age or to the electoral register. A considerable sector of non-voters usually explain their absence from the polls with such reasons as work, duties, unexpected situations, trips, being out of town and illness. Although it is difficult to determine them, in many cases these explanations are really justifications of a more or less voluntary behaviour. In the second place, age problems usually affect the young people and especially the elder. In spite of the young people's usual non-voting (25), the unvoluntary non-voting is usually applied to the most aged electors who are susceptible of suffering physical or psychic problems that can prevent them from voting. The importance of this fact is better seen if we consider the trend towards. a progressive aging of the Western societies. In Spain the rate of people older than 60 among the adult population has rised from 18.2% in 1960 to 23.8% in 1986; a conservative estimate for 1995 would reach 24.3% (26). According to data from different polls, about 15% of them usually do not vote adducing illness in a considerable proportion. We can suppose anyway that the rate is really higher given the general lack of involvement of the elder and of their lesser accessibility to be interviewed (27). Lastly, the large number of people who assert that they could not vote for census' problems is outstanding: between a fourth and a fifth of non-voters in the legislative elections of 1979, 1982 and 1986, according to the respective post-electoral surveys (28). Is it a weighty argument or, on the contrary, just an excuse? Possibly both, although in different degrees for different non-voters. Some studies have outlined the importance of a census created by the Administration, to the point of considering it as a necessary condition -although not sufficientto provide a higher electoral participation (29). But it can also happen that the census affects notably the non-voting levels' distortion (almost always by rising it). Thus the technical non-voting appears as formed by a group of different factors which imply from the census' difficulties to reflect faithfully the population's natural movements until the existence of accumulated mistakes due to its unavoidable aging. It is impossible to make conjectures about the impact of census' imperfections on non-voting, but an error of three or four percentage points has been estimated as probable in many countries (30). The reckonings of the technical non-voting significance have allowed the correction of, the official nonvoting figures -also known as "apparent"- with more adjusted, or "real", ones in countries as the West Germany, France, Italy and the United Kingdom (31). In Spain many analysts have criticized the electoral census deficiencies, which are claimed to be responsible of the non-voting quantitative excess in several percentage points (32). These criticisms have increased recently despite the elaboration of a new electoral census on the basis of the population census of 1980 and its correction in 1986. Thus, for instance, a conservative estimate of the accumulated errors of the 1986 electoral census asserts that the non-voting official figures should, be reduced at least between three and seven percentage points (33). If it is true, as it has been said, that, ceteris paribus, nonvoting will be so much lower as more recent is the electoral census' compilation (34), it seems also true that in Spain the technical non-voting has been considerable even in the moments after the electoral census' creation, getting worse as time goes by. Non-voting mobility Lastly, we must make a short reference to the non-voters' mobility. This is Table6 Table7 Table8 Table9 Table10 Table11 Table12 Table13 Figure1 Figure2