The youth vote in Greece 1990-2023

Panagiotis Koustenis
PhD Political Science

Executive Summary

From the early years of what we call “Metapolitefsi” in Greece -meaning the historical period that came after the 1967-1974 military dictatorship-, the youth played a key role in the political scene, especially after the electoral victory of PASOK (Panhellenic Socialist Movement) in 1981 and in the aftermath of young people’s left-wing radicalisation, which had peaked during the last years of the dictatorship. However, this political assessment was soon reversed in the 1980s, as the gradual conservatisation of the younger age groups became a key factor in the shift in electoral correlations in 1990, which led to the return of Nea Dimokratia (ND) to power.

Since then, the predominant impression was of a rather politically disengaged young generation, something that was in line with similar trends in the Western world at the time. However, in the following years, the electoral choices of the younger generation were reversed yet again, shaping a pattern that was actually the opposite of the general electoral trends, setting the younger generation almost on the sidelines of political and, above all, electoral competition for quite a while.

However, the generalised shift away from the traditional two-partyism that was observed in the late 2000s amongst all voters under the age of 55, was a forerunner of a broader crisis of the established political system, which was to peak in the 2012 electoral earthquake, of which the youth emerged as perhaps the most dynamic pillar. Since then, the majority of the younger age groups (particularly women) have been associated with the SYRIZA party (Coalition of the Radical Left), which brought them back to the forefront of electoral competition, thus leading, at that stage, to a total electoral realignment.

Even when Nea Dimokratia won the elections in 2019 and returned to government, this did not seem to reverse the electoral correlations amongst people aged up to 34, who for the first time seemed to be electorally autonomous from the rest of the voters, a fact that kept them in the electoral spotlight for the following years.
However, the 2023 elections caused a new “small” earthquake, this time mainly referring to the collapse of SYRIZA, while younger voters, in particular, strengthened new challenger parties, in a way confirming the instability that often describes their electoral behaviour, although maintaining some of its more general structural features, such as reduced electoral participation, a more indecisive attitude and, above all, lower party identification and less ideological vote criteria.

Besides, even within the context of their volatile final preferences, younger voters seem to maintain certain standards for the time being, such as the somewhat limited (although still majoritarian and partly balanced compared to the past) appeal of Nea Dimokratia, thus still constituting the least privileged audience for the government party. Voters under 35 act uniformly and to a considerable extent distinctly, as a generation defined by the experience of the economic crisis and its consequences. However, it emerges that these younger electorates remain more susceptible to choices that may seem to be unpredictable at first, especially at a time when post-democracy seems to be gaining ground.

Introduction

The significance of conducting voting pattern analysis based on people’s age is based on the assumption or the conclusion that the very stage of life that the voters are in, is a crucial variable that determines, first of all, their political and then, their electoral behaviour. Since the early years of the field of electoral sociology, various theories or observations have been put forward, often in the form of almost stereotypical norms about the relationship between age and voting, such as, for example, that younger voters tend to be more left-wing or “progressive” while older voters, conversely, are usually more conservative, a belief that was apparently reinforced in the wake of the radicalisation of youth in the 1960s.

This observation, despite its relatively consistent verifiability in the course of time, has occasionally fluctuated and in some cases has even been questioned, especially after the 1980s and 1990s¹. Since then, young people’s electoral behaviour (despite any internal variations) has become more associated with traits of political detachment, such as reduced electoral participation, more de-ideologised or individual-centred (“rational”) criteria, lower party identification etc., while any occasional inclination towards the Left is often directed towards its more alternative expressions². These changes in relation to the past, actually apply to the entire electorate, but in the case of young people, they are much more pronounced, a fact which distances them from their earlier association with the Left and, more importantly, from its traditional values.

Therefore, it becomes clear that any distinct behaviour of younger voters in particular, is directly related to the specific historical context, social perceptions or the general “political climate” of the era when they start to socialise politically, which, in turn, is often determined by the moment of coming of age and acquiring the right to vote and is considered to exert a long-term influence on their choices, thus giving each separate age category the characteristics of a “generation”³. Of course, this perspective presupposes a relative consistency of said choices over time, which is not always the case for the younger generations, at least until the moment of their social integration.

However, even in this unstable situation, the concept of generations and especially that of “youth” is highly significant in explaining major upheavals or realignments in the electoral landscape and the broader political system⁴, when these emerge either abruptly in the context of a single “critical election”, characterised by a radical and structural change in the previously established divisions of political competition, or gradually, through a long series of electoral processes (secular realignment)⁵. In theory, such shifts can be manifested or initially interpreted as intergenerational gaps, with younger generations acting as heralds for those changes, since they are less “bound” by traditional electoral identifications⁶. When a similar behaviour is adopted to a certain extent by the intermediate age groups, then a total reversal of the political scene can be assumed to happen in just a matter of time, as it is expected to occur with the gradual age replacement within the electorate, i.e. with the departure of the older age groups from it and the subsequent change in its overall composition⁷.

Age-related differences in voting behaviours during the years of Metapolitefsi

In public opinion surveys, the behaviour and attitudes of young people, in the most “narrow” sense of the term, are first of all assessed from data concerning the 17-24 age group (i.e. those at the threshold of entry into the labour market). However, this approach is very often extended to the entire general category of 17-34 year olds, a practice that also increases its statistical significance in terms of the overall size of the adult population and, consequently, the electorate. Indeed, in recent years the second (extended) approach has proved to be increasingly useful, especially for the Greek example, as it covers a more comprehensive range of age groups that have experienced the consequences of the preceding more than ten years of economic crisis, while the individual electoral differences (above and below the age of 25), at least in terms of the influence of the major parties, are becoming systematically smaller compared to the respective records before 2009.

In Greece, due to the delayed development of analytical political research, up until the first years of the Metapolitefsi, any indications of the political or electoral behaviour of younger people were primarily based on secondary data (e.g. the results of student elections) or on assumptions based on each period’s general political climate. For example, the youth of the 1960s were considered to be radical and left-leaning, as was the generation of the student uprising of the Polytechnic School in Athens, the first to take centre stage in the political arena with the rise of PASOK to power in 1981⁸.

However, this situation began to change and was eventually reversed during the 1980s, in line with the broader political changes in the international arena, as the first systematic opinion surveys of the time revealed. This tendency culminated in the elections of 1989-1990, when a clear over-representation of Nea Dimokratia (ND) and, secondarily, of the Green Party⁹ among the youth was recorded, thus placing the Greek case among the exceptions to the stereotypical “norm” of left-wing youth and ultimately playing a crucial role in the reversal of the overall electoral correlations at that particular juncture.

From then on, the impression of a more conservative or de-ideologised young generation began to prevail, in line with similar trends that then prevailed in the Western world¹⁰.

1993-2009

This pattern was confirmed in the 1990s, with the introduction of exit polls in Greek political analysis as a fundamental tool that allows us to analyse people’s voting behaviour¹¹.

Indeed, as far as the electoral competition of the two-party system is concerned, the overrepresentation of Nea Dimokratia amongst the younger age groups (especially those aged 18-24) is already reflected in the 1993 elections (based on the exit poll conducted in the European elections of the following year), despite the fact that the most privileged age group for Nea Dimokratia was still that composed of voters aged 55 and over (Table 1). This was not just an undisputed fact, but also a trait that was reinforced in the following years, at a time when PASOK’s over-representation was mostly found in the intermediate age groups (35-54 years old).

However, at the same time, a reverse process seems to be underway. Although Nea Dimokratia maintained its dominant share amongst the youngest age group (18-24 year-olds) at rates higher than or equal to its national average until 2000, this share was constantly decreasing, as PASOK’s influence in the same age group steadily increased after the two successive changes in government (in 1996 and 2004) as well as with the gradual entry of a new generation into the electorate during its second term in office. The result was a definitive shift in the electoral correlations amongst the youth from 2004 onwards, as well as the overall restoration of their predominantly left-wing (or rather, “anti-right”) positioning, although not with the same comparative intensity as in the past, since at that time, the Left as such (namely the KKE and Synaspismos) no longer had its past dynamic amongst young people.

As a side note, it should be pointed out that the aforementioned development works completely in reverse from the final outcome of the elections, a phenomenon that is even more clearly reflected in the student vote (according to people’s relevant professional status declaration when participating in exit polls), where Nea Dimokratia consistently prevailed until 2000, when they were the main opposition, while they lost their majority immediately after their return to power in 2004.

In other words, from 1993 until 2007, the party that won the majority of the (student) youth votes was ultimately the one that lost the elections (Table 2a and 2b), a fact that suggests that the youth vote does not always coincide with the broader electoral context, nor is it in itself capable of leading to major political changes, when it is simply registered as an exception or as an isolated reaction and is not followed by a more widespread electoral trend that also crosses other (usually intermediate) age categories.

Nevertheless, the 2004 elections proved to be an overall turning point in terms of the distribution of the votes according to age and the political parties’ corresponding profiles. Nea Dimokratia’s victory that year was primarily the result of a 5 percentage points increase in the party’s appeal to older voters over 55 (from 47% to 52%, see Table 1), or even of 7 percentage points, amongst voters over 65, while its share of the youth vote remained the same since 2000 (43%), i.e. there is a gap of about 10 percentage points, with an age cut-off at 55 years of age, which was to be maintained at almost the same level until 2009.

On the other hand, although PASOK lost the 2004 elections, it temporarily (perhaps since the early 1980s) put forward a comparatively more youthful profile, which, however, was not to last for long. The loss of almost 3-7 points exclusively in the under-55 age groups in 2007, was only partially regained when the party returned to power in 2009. By contrast, the key factor in PASOK’s last electoral victory, was a rise in its share of the vote strictly localised amongst the over-55s, by an aggregate of 9% compared to 2004, which completely reverses the short-term “youthful” result of 2004 and also creates an age differentiation similar to that of Nea Dimokratia. Besides, it was precisely at that time (2007-2009) that KKE and Synaspismos (as well as the newly emerging Green Party) started to regain significant momentum amongst 18-24 year-olds, while at the same time, there was a notable over-representation of the 25-34 age group amongst the supporters of G. Karatzaferis’ political party (LAOS)¹².

Ultimately, the result was creating a numerically unprecedented overall 13-point gap in the aggregate share of the two-party voters between those over and those under 55 years old (73% vs. 86% respectively, Table 1). This element, along with the corresponding geographical differentiation between the Attica region and the periphery, which has been consolidated since 2007, were perhaps the two main cracks in the overall countenance of the Greek electorate in the 2004-2009 period and the earliest precursors of the crisis that would follow.

The recession years: 2012-2019

Indeed, this phenomenon took on explosive proportions, initially in the electoral earthquake of May 2012, where the aggregate share of the two main political parties of the Metapolitefsi era plummeted to 20%-25% amongst 18-54 year-olds (with relatively small partial fluctuations), compared to 49% amongst older voters (or even 58% in the case of the over-65s).

Thus, the age differentiation of the vote was determined as one of the three basic rifts (along with the geographical and the social), which essentially divided the electorate into subgroups, almost unconnected to one another. The mass abandonment of the traditional ruling parties by the younger generations and their turn to (mainly anti-memorandum) then emerging parties was one of the most vivid manifestations of the Memorandum-Anti-memorandum divide, which in this case was condensed in a distinction between the “old” and “new” political systems¹³. This pattern was more or less reproduced in the subsequent elections in June, where Nea Dimokratia and SYRIZA increased their shares by approximately 10%. This led to the emergence of a new two-party system, stabilising the age threshold for each party’s respective majorities at 55 years (Table 3).

In this sense, the youth in this particular situation did not display a wholly distinct electoral behaviour, but rather symbolically constituted a kind of “vanguard” of a more general pattern for all voters under 55, leading to an overall electoral realignment. A fact that was nonetheless to reassert its electoral role in a broader sense in the years to come, not simply as a pool for renewing the parties’ electoral base, but again as a key factor determining electoral competition.

Nevertheless, this double election also marks the beginning of the (initially erroneous) perception of SYRIZA as a mainly “youthful” party, when its highest net percentages were recorded in the dynamic age groups of 35-54 year-olds, which mostly consist of people active in the labour market, in other words, the main pillar of its rise to power in 2015.

The confusion was likely due not so much to SYRIZA’s high figures in the under-34 age group (30% in the June elections), but mainly to the fact that said numbers not only granted the party a larger majority (Nea Dimokratia saw its share of the vote amongst the same age group drop dramatically – 22%), but even exceeded the corresponding aggregate of the old two-party system (28%). At this point, we should stress the fact that other parties that emerged in the context of the economic crisis had a much “younger” footprint, with their percentages being clearly age-graded. The most significant and long-lasting example, until 2019, are the percentages of Golden Dawn¹⁴, and to a lesser extent, especially for 2012, those of ANEL and the Green Party, while in 2015, a similar picture was observed in the case of Enosi Kentroon and LAE. In the same period, the exact reverse age distribution was consolidated for the PASOK electorate, whose percentages in the age group of 55+ were consistently more than double those of voters aged 18-34, thus making it the most “aged” party in electoral terms¹⁵

For SYRIZA, in the elections that took place in January 2015, the main difference compared to 2012, was the growing trend of demographic – and geographical – homogenisation (which had already been discernible since the 2014 European elections). SYRIZA’s reach increased sharply, primarily in the 65+ age group (by 17%) and then also in the 55-64 age group (by 15%), and in fact, in the latter it recorded one of its highest individual performances (39%), taking the lead from ND and finally raising the age limit of the electoral competition between the two parties to 65, a development that secured SYRIZA’s final victory in the elections.

The main characteristics of this age distribution were essentially repeated in September 2015, with the key difference this time being the further increase of SYRIZA’s appeal amongst 18-24-year-olds (and the simultaneous respective drop of that of ND), with the party recording its strongest ever percentage (42%), in the aftermath of the July Referendum, where this particular age group – more than any other – had voted for NO (above 75%)¹⁶. This view of SYRIZA, as a truly “youthful” party, was then reflected even more vividly in the 2019 elections, which also marked the end of the Memorandum years. Its share in the 17-34 age group was maintained at the same level as in 2015 (37% overall, despite an interim drop to 22% in the previous month’s European elections).

At the same time, Nea Dimokratia’s corresponding share was 30%, while amongst voters over the age of 65, it was almost 50%, yet balanced compared to September 2015. However, the most critical change for the final outcome of those elections has to be the drop in SYRIZA’s share among the dynamic age groups (35-54 years old, by 7% in total) and the conversely significant strengthening of Nea Dimokratia (+14%), with the dividing line between the two parties descending to 35, reflecting the total reversal of the electoral correlations in the core of Greek society¹⁷.

It was perhaps the first time that the youth as a whole (up to 34 years old) was seen as an autonomous electoral social-demographic group, compared to the rest of the voters, a fact that naturally brought it to the centre of electoral interest, with relevant discussions intensifying dramatically just prior to the last elections, after the tragic train accident in Tempi and the mass protests that followed. Furthermore, it is worth mentioning the timing of the 2023 elections that took place, just like in 2019, at the end of the four-year term, that is when four young age groups (i.e. more than 400,000 new registered voters – for the second consecutive time), would join the electorate.

However, in order to determine the real impact of the younger age groups’ vote, one should also take into account the continuous decline in its quantitative value, due to the demographic ageing (plus, to a certain extent, the brain drain phenomenon) observed in recent years. While the younger age bracket (18-34 years old) in the 2001 census represented 32% of the country’s adult resident population, this percentage decreased to 27% in 2011 and is estimated to be around 22% after the most recent 2021 census (with the addition of the 17-year-olds), when the corresponding percentage for voters over the age of 55 is about twice as high.

This demographic development, largely verified by the respective participation rates of each age group in the exit polls, makes it much more critical for a party such as Nea Dimokratia to maintain its lead in the older age groups, as the leading opposition party, currently SYRIZA, needs twice as big a share in the younger age groups in order to compensate for it.

The double elections of 2023

As with other key dimensions of the vote, namely its social and geographical aspects, the most defining trait of the recent double elections age aspect was the confirmation (and in some cases even the strengthening) of Nea Dimokratia’s absolute dominance (40.8% and 40.6% in both elections) and, on the other hand, not just SYRIZA’s dramatic electoral shrinkage, but also the almost complete flattening in terms of its influence. Indeed, all its individual percentages by age category are, with some deviations, around its overall national average (20.1% in May and 17.8% in June), bearing little resemblance to the former age distribution of its electoral base.

It is indicative that, with minor variations, SYRIZA’s comparatively lowest percentages in both elections were recorded amongst the intermediate age groups of 35-54 year-olds (Table 4), i.e. amongst those who used to be its most crucial and once privileged dynamic audiences, and where it had suffered the greatest damage in the 2019 elections. At the same time, in the 17-34 age group, the party lost an overall 17 percentage points, i.e. almost half of its former electoral power. This fact dashed any expectations that it could be an “antisystemic” expression of the youth movement after the Tempi train accident.

On the other hand, the age distribution of Nea Dimokratia’s electoral influence has remained essentially unchanged since 2019, basically replicating the same pattern: around 30% amongst younger voters (17-34 years old), 35%-45% amongst intermediate ages (35-64 years old) and 50% or higher amongst older voters (over 65 years old). The data do not verify the often prevailing impression of a strengthening of ND’s influence on young people, which is created due to the advantage it gained as a result of the complete collapse of SYRIZA’s figures. It is also worth mentioning the relative levelling out of the age differentiation when it comes to the PASOK vote in the last elections (10%-11% amongst voters under 55, against 13-14% amongst those aged 55+), thus essentially getting rid of the impression of a political party with an ageing electoral base that prevailed throughout the previous decade.

So, looking into their voters’ age distribution in particular, SYRIZA and PASOK went from presenting two entirely complementary profiles, up until recently, to a more competitive situation, which nevertheless still favours SYRIZA. Finally, the increase of KKE’s percentage was significant, with the 17-34 age group being its comparatively strongest supporter (8%-9%) in both elections, a fact that is relatively unusual for this particular political party since the mid-1980s, but which points to the two consecutive victories of their main student branch “Panspoudastiki” in the past two student elections.

The aggregate percentages of these four traditional parliamentary parties, amongst the three broad age categories (17-34, 35-54, 55+) were 73%, 80%, 86% in May and 67%, 74%, 86% in June, with the gradation of the relevant percentages resembling a miniature version of the electoral earthquake of 2012 or its precursor symptoms of 2007-2009.

However, on this occasion the respective gradation is almost exclusively determined by the corresponding escalation of the rates of ND, since the respective fluctuations for the other three parties are more contained. It thus becomes obvious that any dissatisfaction or the (often misleadingly called) “anti-systemic” nature of the youth vote was expressed mainly at SYRIZA’s expense and in favour of smaller players in the electoral competition, i.e. the parties that sought their entry (or their stay) in parliament, with the respective aggregates of their percentages amounting to 27%, 20%, 14% in the first elections and 33%, 26%, 14% in the second. In other words, younger audiences voted for such parties at levels twice as high as those of older voters in the May elections and this pattern was further boosted in June, taking into account the increased abstention by 800,000 votes (or 700,000 valid votes).

MeRA25 had the proportionally youngest vote by far in the last double elections (as it did in 2019 as well), with its share remaining around 6% in the younger age groups (up to 34 years old), while slightly decreasing to 2%-3% in the older age groups, from 2%-4% respectively in 2019. However, this change, coupled with the aforementioned current demographics, probably cost the party its participation in Parliament. Plefsi Eleftherias had a similar, but slightly more balanced distribution, receiving 5% in the younger age groups (6%-7% amongst 17-24-year-olds) and 2%-3% in the others, but in the second elections the party’s share was marginally boosted, just enough to finally cross the 3% threshold.

Meanwhile, Spartiates’ influence, as recorded in the June election, was very much reminiscent of Golden Dawn’s youth vote in the past, with a greater uniformity in all categories under the age of 45 (8%), yet clearly with less than half of GD’s percentage in the older age groups (2%-4%). Finally, the overall vote for Other parties reveals a similar youth distribution, despite its shrinkage in the June election, while the data for Elliniki Lisi and Niki do not follow the same pattern, with the former appearing more balanced and the latter having more impact on the intermediate age groups.

At the same time, one can highlight the internal divisions of the youth vote, by cross-tabulating it with gender and educational level (Table 5), however, in this case, the findings generally confirm and often magnify the corresponding differentiations recorded in the electorate as a whole. More specifically, there are more younger male voters (17-34) who support and vote for Elliniki Lisi (5%-7% compared to 3%-4% amongst the general population) and this is even more prominent in the case of the young male Spartiates’ voters (12% vs. 3%), a finding that is perhaps the most recognisable common element with Golden Dawn’s past demographic profile.

On the other hand, the female voters favour Plefsi Eleftherias (4% to 6%-7%), PASOK (8% to 12% in June) and mostly SYRIZA, especially after its share dropped to 16% among young men in June. In fact, young women were perhaps the only subset of the electorate in which SYRIZA’s share was comparable to that of ND. A contrast which directly points to the consistently observed more left-leaning vote of young women in recent years¹⁸, something which is not confirmed in the case of MPA 25, especially in the second elections (7% amongst young men against 4% amongst women).

With regard to the differentiation based on the (completed) education level, between secondary and tertiary education, meaning having or not a university degree (applied only for the 25-34 age group, where it makes the most sense, statistically), apart from the traditional parties (and with the exception of PASOK in May), the highest share of young people with higher education (university degree) are observed amongst the voters of Plefsi Eleftherias (in June) and mainly those of MeRA 25 (just like in 2019). On the complete opposite side, Elliniki Lisi (8%-11% to 2%-3%) and Spartiates (13% to 5%), i.e. the two main parties positioned at the rightmost extreme of the traditional ideological axis, have significantly higher percentages among young secondary school graduates (no university degree), which points to the mainly cultural significance of the education variable amongst the younger age groups (given the fairly high ratio of graduates from higher education institutions), as opposed to the older ones, where its value still has a socially divisive dimension, but which is now mainly limited to voters over 65.

Although the latter findings were perhaps to be expected, there are, nevertheless, two significant differences that emerge at this point between the recent elections and those of the 2012 “electoral earthquake”. The migration of a large share of the voters to newer challenger parties in 2012 was a practice observed mostly amongst people with higher education, while this time, to a comparatively greater extent, it concerned voters with a lower education level, especially among the younger age groups.

Still, overall, the younger age group’s voting behaviour continues to be predominantly left-leaning (or rather, “non-right”), although with less intensity than in the past, but still in contrast to the rest of the age groups, as well as to all the other demographic-social groups that constitute the electorate. This indication is obtained prima facie by adding the percentages of the “non-right-wing” forces in the actual election results: In both elections that sum is approximately 50%-52% (compared to 41%-44% in the case of all voters over 35), if we include in this calculation the percentage of PASOK and that of the (ideologically more controversial) Plefsi Eleftherias.

The exit poll responses on the voters’ ideological self-positioning on the Left-Right axis lead to similar conclusions. Quantifying them and assigning values 1-7 on said scale (where 1 = Far Left, 7 = Far Right), the average for the 17-34 age group is lower (that is, closer to the Left) than the median value of 4 (that marks the Centre) in both 2023 elections (3.82 and 3.96 respectively), despite its overall shift towards the Right in the June elections (Table 6d), with a noticeable distance from the 35-54 and 55+ age groups (which are to the right of the Centre, so at values higher than 4), a distance significantly greater than that recorded in similar measurements in the past¹⁹.

The above is based on “valid” responses, i.e. those of perhaps the most politicised participants. At the same time, however, it is worth mentioning the younger age groups’ much higher overall rates (34%-36%) of rejection or non-acceptance of the ideological axis (in other words, the “None of the above” option), compared to the intermediate age groups of 35-54 year-olds (26%-28%) and the older age groups (55+:14%-16%). Thus, the general perception of the youth’s increasingly de-ideologised behaviour is confirmed, which to a certain extent also conceals “anti-political”-possibly far-right or other socially “unacceptable”- views, which coexist with the more Left-leaning stance of the more openly politicised part of the youth, thus leaving room for unexpected electoral outcomes.

Indications from further processing of the exit poll data seem to confirm additional characteristics concerning the general voting behaviour of younger voters. For instance, they appear to be more undecided: 42% in total decided who to vote for in the last week or on the day of the elections in May, compared to 30% and 16% among the other two age groups (in June the respective percentages were 20%, 16%, and 8%). They also show significantly lower levels of identification with the party they end up voting for (42%-47% vs. 55%-63% for the entire electorate) and are less likely to cast a positive “vote of preference” (44% vs. 49% and 62% for the other two age groups in May). Finally, based on the participation rates of each age group in the exit polls of the two consecutive elections, the increased abstention in June seems to concern younger voters to a much greater extent, as their turnout is estimated to have decreased by 25% compared to May, when the corresponding decrease for voters aged 35-54 was 17%, and the participation of older voters (55+) seems to have remained at almost the same level (-2%).

Electoral background by age group

Coming back to the concept of generations, we attempted an investigation regarding the temporality of the voting behaviour of each individual age subset of the electorate, depending on the exact year of joining the electorate in parliamentary elections (taking into account the three voting age expansions in 1977, 1984 and 2016, with the age limit changing to 20, 18, and 17 years old respectively, instead of 21 years old). This process is rendered possible through the use of recent exit poll data (from 2012) and checking the ages of voters as calculated from their exact year of birth (a question usually answered by 95%-97% of all participants).

The biggest methodological problem with such an approach is dividing the total sample into tiny groups²⁰, thereby significantly increasing the statistical error. However, the aggregate size of the exit poll (samples of more than 5,000 people), which is far larger than that of almost all similar social or political surveys, ensures relative statistical reliability (similar to that for large age groups in the usual political surveys of 1,000 or 2,000 people), even for these “narrow” age subsets, especially with regard to estimating the electoral influence of the major parties.

Based on the above, Table 7 contains an indicative overview of ND’s intertemporal percentages per age group, from 2012 until May 2023 (the June elections are not taken into account in principle, due to the extended abstention rate, which is partly responsible for any differentiations).

The groups now aged 17-36 years (born in 1987-2006, average sample: 244 people in May 2023), i.e. those who became eligible to vote after 2005, are basically the generation of the economic and political crisis, whose politicisation probably began with the events following the assassination of Al. Grigoropoulos or slightly earlier.

These voters to this day are still comparatively the least supportive of ND, as even in the 2019 elections, this group’s votes for ND barely exceeded the 30% mark (indeed, in September 2015 they were well below 20%), something that happened in the elections of May 2023, even though briefly, since in the context of the generalised abstention in June, the numbers dropped again. Conversely, the above age categories, up until 2019, were all consistently giving SYRIZA the majority (the relevant figures were usually 35%-40% in the two 2015 elections), but this changed in 2023, when they left the party in second place (20%-25% or even lower in June).

Those born between 1973 and 1986, now aged between 37 and 50 (average sample size per group: 363 persons in May 2023), constitute the group that became eligible to vote in one of the most de-ideologised periods of the Metapolitefsi (1993-2004), given the convergence or consensus between ND and PASOK on issues mainly related to European strategy. Among these voters, ND did not exceed 28% until 2015, but in 2019 it got perhaps the most impressive boost in its percentages (by 16%-20%), overall displaying the highest electoral mobility within the electorate.

However, in the 2023 elections, there was an internal division in this category, as amongst the older of these voters (those aged between 45 and 50 today), ND’s performance increased a bit more, while amongst the younger ones (aged 37-44) it suffered a decline of around 5%. That is, the latter ended up being closer to the patterns of the people belonging in the immediately younger group and the former to those of voters in the older category. In this way, however, a clear and critical dividing line was formed, perhaps more explicitly than ever before, regarding ND’s influence, with 45 years of age as the limit, not only for this particular age group, but also for the entire electorate as ND’s average percentage is estimated at 46% amongst all voters over the age of 45, and 32% amongst those up to 44 years old.

Voters aged 51-69 (born in 1954-1972, average sample size per group: 481 people in May 2023) are those who came of age during a period of intense politicisation (1975-1990). Within this group, ND shows a more gradual, but still continuing, expansion of its reach: starting from 25%-30% after the June 2012 elections, when it was on second place behind SYRIZA, and even after a temporary decline of 5% in January 2015, its influence rose to 36%-39% in 2019, but almost never exceeding its national percentage. However, its new rise to 42%-44% in 2023, if it does not prove to be conjunctural, may maintain an expanded potential electoral base in the years to come.

On the contrary, the next oldest age group, the 70+ individuals (born until 1953, total sample in May: 835 people who became eligible to vote by 1974), the majority of whom will not be active in the electorate at the end of this decade. This is the group in which ND has always recorded its highest percentages by far, with its share reaching 40% (on average) from June 2012 until 2015 (it was 33% even during the May 2012 electoral earthquake), until approaching and subsequently exceeding 50% in 2019 and 2023. Of course, this age group can also be subject to age and historical distinctions, e.g. between those who came of age during the dictatorship (70-76 years old today), and voted for the first time in 1974, and older people, i.e. those who would be able to vote in 1967, had there been elections instead of the military coup.

The data presented in Table 7 show that for the first group (the “dictatorship generation”), with all its particular historical and political traits, at least as far as ND’s appeal is concerned, while in 2012 its percentages were much closer to those of the younger age groups, gradually, after January 2015, the relative distance from the older age groups began to shrink. This could be an indication that after a certain threshold (possibly between the ages of 55 and 65), age per se is a more significant determinant of voting behaviour than the generation in which a person “belongs”.

In any case, the above analysis is essentially just a first approach to the temporality of voting patterns by generation, open to multiple additions or revisions in future papers, but indicating a technical and methodological possibility. However, on the one hand it demonstrates that the usual age groupings in polls are likely to obscure the reality of age differentiation regarding voting, to some extent, while a new grouping, following a detailed listing of the individual data, might be more illuminating. An additional element supporting this suggestion is the observation that the differences in percentages between close or successive age groups are relatively limited, often giving the impression of a certain homogeneity. When this is not the case, then the age divisions within the electorate become more visible.

To conclude, however, with regard to the data presented in this paper, which concern exclusively the study of ND’s influence in this particular time period, we should mention that while in some cases there appears to be an overall homogeneity of the vote in almost all age groups below 55-65 years old (see May 2012 and January 2015), in others the existence of a broader internal division (sometimes more powerful and clear, and sometimes not) of the more dynamic audiences becomes apparent, with the relevant threshold usually ranging between the ages of 35 and 45 and with the possibility of perhaps more individual groupings.

Obviously, the youth, by which I mean all voters under the age of 35 or so today, is included in these as a distinct and relatively uniform age group now, consistently emerging as the least favourable audience for ND since the 2012 electoral earthquake and confirming its electoral behaviour with the characteristic traits of a “generation”, that has been through the common experience of the economic crisis and its consequences, and feel the same political disdain for the two-party system in all its forms. Similar intertemporal historicity may be detectable in other sub-age categories, but to strengthen such conclusions it would be necessary to extend the intertemporal overview to the electoral influence of the other parties in future research.