Lowering the Voting Age to 16 in Practice: Processes and Outcomes Compared

Research into the possible consequences of lowering the voting age to 16 used to be rather speculative in nature, as there were few countries that had implemented earlier enfranchisement. This has changed over the past decade. We now have a range of countries in different locations, mostly in Europe and South America, where 16- and 17-year-olds can vote in some or all elections. In many of those places empirical research has given us insights into the experiences of young people and the impact of those changes on political discussions. However, so far these studies have largely been conducted individually in each country, which makes comparisons difﬁcult. This article summarises the key insights from empirical research across countries with lower voting ages. It identiﬁes common patterns, but also highlights differences. Overall, the impact appears to not be negative and often positive in terms of political engagement and civic attitudes. However, the comprehensiveness of effects varies. The article offers some possible frameworks to understand differences, in particular by reﬂecting on the processes that led to voting franchise changes, but also indicates where gaps in knowledge remain, and what sort of research would be required to produce sys-tematically comparable results.


Introduction
Since discussions about lowering the voting age emerged in the early 2000s in several countries, much has been written about what we should expect if the franchise was to be extended to include 16-and 17-year-olds. Critical contributions to the debate ranged from the more polemical (Russell, 2014) to research analyses (Chan and Clayton, 2006) and government sponsored consultations (see, e.g. Youth Citizenship Commission, 2009). Contributions that were favourable of lowering the voting age often came from a more advocacy-based standpoint (see, e.g. campaigns by the British Youth Council (2020) or the Electoral Reform Society (2020)). Amongst the sceptical accounts, many of the analyses suggested that the concerns outweighed the potential positives. Crucially though, most of their arguments relied heavily on inquiries that tried to establish what might happen, rather than an examination of what actually occurs when 16-and 17-yearolds are allowed to vote. However, as has been pointed out elsewhere (Wagner et al. 2012), there are significant problems with this approach. Typically, those studies looked at slightly older young people, such as the age group 18-24 and showed that indeed those young people tended to be less politically engaged than the average population (Johnson and Marshall, 2004). Those findings were then extrapolated to the even younger age group, assuming that any negatives found would be extended to them, too. However, that sort of approach does not allow for voting at 16 or 17 to be a different experience. Researchers began to ask what, if rather than doing the same thing a bit earlier, being enfranchised at a younger age actually changes what the first engagement with representative politics looks like?
Together with the desire of scholars to deepen research on the topic, data availability on actual empirical experiences is an important prerequisite to answer that question. Studying it is not possible, unless we have places in which 16-and 17-year-olds are allowed to vote. Because of that, debates in many countries indeed largely relied on rather speculative investigations until recently when several countries began to reduce their respective voting ages below 18. Experiences with voting at 16, however, are not entirely new. A small number of countries did this in the 20th century already, such as Brazil in 1988 (following earlier changes in Cuba in 1976 andNicaragua in 1984). However, little attention was paid to those experiences in the abovementioned studies published in the early 2000s in European countries. A second wave of franchise extensions changed that. Starting in the mid-1990s, some states (Länder) within Germany began to lower the voting age for municipal and later state-wide elections, the same process began in some parts of Austria in 2000. The magnitude of change increased, however, when Austria in 2007 extended the new voting age country-wide and for all levels of elections, including those at country level, thus joining the countries that had done this already three decades earlier. In 2008, Ecuador decided to do the same. This provided new opportunities for empirical studies into what impact a lowered voting age could actually have in practice. Opportunities for new data collection continued to grow throughout the following decade. More German states allowed 16-year-olds to vote, Norway ran two waves of experimental studies at the municipal level in 2011 and 2015 and Argentina lowered the voting age in 2012 at the national level. Scotland let younger voters take part in the 2014 independence referendum, before extending the franchise for all Scottish elections in 2015. Estonia joined in for local elections in 2015 and in the USA, we have seen some successful and some unsuccessful local initiatives to lower the voting age in certain municipalities. The process continues with Malta having lowered their voting age in 2018, Wales doing the same for the 2021 Welsh Parliament elections and debates being held in many other places.
Crucially, there is now no excuse anymore to simply speculate about what would happen, if 16-and 17-year-olds got to take part in elections. We have a wide range of countries in which they can and we have data from many of those countries, providing us with insights into the behaviour and attitudes of those newly enfranchised people. While the context in which those young people experience taking part in politics is, of course, specific to the particular country and contingent on the character of the political system, bringing together the insights from across different countries on this topic is very meaningful and an endeavour that an increasing number of scholars have undertaken. Being able to examine what really happens when 16-and 17-year-olds are allowed to vote, this article aims to summarise what we know (and do not know yet) about earlier enfranchisement. We compare empirical studies from countries that have lowered the voting age and discuss what similarities and differences we are able to observe.
We begin by looking at the different processes that lead to the lowering of the voting age and discuss why the process of enfranchisement matters. This allows us to show how being able to vote at 16 and 17 is not simply an earlier version of what would have happened a few years later, but instead something that can impact young people in a country. Following on from this, we examine what this impact actually looks like in practice. Using the findings from empirical studies across countries, first we look at young people's political behaviour (in particular their participation in elections), followed by a discussion about their political and civic attitudes and finally potential impacts on the broader debate about young people and politics more widely. While some findings are rather uniform, others differ or are highly moderated by specific contextual factors interacting with the voting experience. We utilise a range of analyses from multiple authors on specific country-case studies brought together in the most comprehensive volume on votes at 16 to date (Eichhorn and Bergh, 2020) and further sources that provide insights into those countries. At the end, we briefly discuss what gaps still exist in our understanding and what we should do to address those in the future.

Top-down or bottom-up? The process leading to earlier enfranchisement
The processes that have led to a lowering of the voting age were not identical across all the countries studied. Even amongst the early adopters in Latin America, there was a significant variation in the political systems of the countries and in the way that the changes came about. While the process in Cuba could be described as top-down, Nicaragua and Brazil saw the movement towards votes at 16 much more embedded in a bottom-up process (Sanhueza Petrarca, 2020). In Cuba, the franchise changes were part of a much broader process of constitutional change that was coordinated and directed by the government. While public discussions and consultations with citizens took place which resulted in several changes to the initial proposals (Nohlen, 2005), fundamentally the process was driven from the top. In Nicaragua and Brazil early enfranchisement was also part of wider constitutional reforms, mainly connected to the democratic transitions the countries were going through. However, the dynamic involved bottom-up elements, in which civil society can be seen as inducing the changes. As, for example, previously marginalised communities in Nicaragua were enabled through a range of programmes to become involved in political processes, young people played an important role through several youth organisations supporting political change (Sanhueza Petrarca, 2020). This was partially recognised in the change to their enfranchisement. Similarly, the transition in Brazil included extensive calls by civil society groups to create social justice for all social groups and enable democratic participation (Lemos, 1988), of which the inclusion of young people formed an important aspect. Lowering the voting age to include them complemented other new measures focussed on direct democracy to develop a deeper democratic system (Sanhueza Petrarca, 2020).
We therefore see, that changes to the voting age are not merely a policy area in its own right, but deeply connected to the broader political context (Mycock et al., 2020). Understanding the process leading to its adoption is therefore an important aspect of examining the impact that it may or may not have.
In the second wave that mostly took hold in Europe and some further Latin American countries, initially, most processes were much more reflective of a topdown approach overall. This is not to say, that civil society groups were not involved in discussions about a lower voting age, but in the majority of cases the change was initiated by incumbent government parties. In some instances, new laws were adopted uniformly across the whole country, often connected with wider reform agendas (e.g. in Ecuador and Argentina, respectively). In other places, regional party groups and later on national parties advocating for a lower voting age could gradually adopt the position in their respective manifestos, after having seen the implementation by their colleagues in power in other regions of the country, such as in Germany or Austria (Aichholzer and Kritzinger, 2020;Leininger and Faas, 2020). While other civil society organisations also supported and even campaigned for the measures, the decision to make votes at 16 part of party platforms and to implement it when in power, were commonly driven by dynamics internal to the party structures.
While top-down dynamics have been more prominent in the early 2000s, the interplay between civil society and central actors within each national partypolitical system can be more complex nevertheless. In Estonia, for example, the idea to extend the franchise to 16-and 17-year-olds was popularised by a major youth organisation in the country (Explanatory Memorandum, 2014), but then largely taken up and developed as a project with ownership by certain actors within the government in a more top-down fashion (Toots and Idnurm, 2020). In the Scottish context, we also saw an interplay of actors. The initial impetus for change was very much top-down and directed by the Scottish Government, run by the Scottish National Party (Huebner and Eichhorn, 2020). However, the proposal also gained support from others, including unionist opposition parties (Labour and the Liberal Democrats). Subsequently, many youth organisations and institutions became involved, including the Scottish Youth Parliament which made votes at 16 a major campaign issue (SYP, 2012). The new franchise, initially only implemented for the independence referendum of 2014, was ultimately adopted for all Scottish elections by the Scottish Parliament in 2015.
As we have seen, debates about the voting age have not taken one singular format. On the one hand this reflects the particular political context within which they emerge. But examining the difference in approaches to franchise changes matters also in terms of the likelihood of official and popular support more widely. As Douglas (2020) shows in his study of municipalities in the USA that saw referenda or representative bodies' votes on lowering the voting age to 16, outcomes of such processes can vary greatly. Successful campaigns usually built on bottom-up approaches that were directed by young people themselves, but also developed significant buy-in from certain political actors. Approaches that were either directed towards, rather than shaped by young people or that lacked representatives' support were less often successful. Bigger, national campaigns on the issue have emerged (see, e.g. Make it 16 (2020) in New Zealand or Vote16USA (2020)) that aim to capitalise on the insights from successful efforts.
In countries where top-down implementation of franchise changes were successful, the issue typically was less politicised overall. However, when debates about the voting age entered the public realm, its fit with discussions in the broader political context and the engagement of civil society actors are of great importance to understand how and why the voting age was lowered in some countries at particular points in time. Crucially, when the issue got debated more extensively, a range of arguments about young people's political behaviour and attitudes and the potential impact on the political system more widely could be found repeatedly-often speculative in nature. In the next section, we will turn to the insights from empirical studies in the countries that lowered their voting age to examine what arguments were found to be supported in practice and which ones were not.

The impact of lowering the voting age
As is clear from even the most casual look at the list of countries in Table 1, the vast majority of countries in the world that hold democratic elections have not lowered the voting age below 18. In a number of these countries, especially in Europe and North America, there is some ongoing debate about the issue. Political parties on the centre-left tend to favour lowering the voting age and may bring it up occasionally. Sometimes, when there is a debate about reforming electoral laws the voting age issue may also come up.
While some of the debates are strictly normative in nature, others lend themselves more easily to research and to empirical testing. The normative debates may deal with definitions of what it means to be a voter in relation to other markers of adulthood (Electoral Commission, 2003a)-discussions about which already marked the process of lowering the voting age from 21 to 18 several decades ago (Loughran et al., 2019). Such debates include, for instance, the question whether voters should be legally and financially independent of their parents or whether youth should be given more political influence through a lower voting age to counter the voting power of a growing elderly population.
When it comes to empirically testable propositions or arguments, the most often referenced issue is that of voter turnout. We know that younger people generally tend to have lower rates of turnout than others, which may make some sceptical of further extending the franchise (Youth Citizenship Commission, 2009). Others argue that giving young people the chance to vote earlier in life will also give them a habit of voting that over time will have a positive effect on turnout levels (Champion, 2014). Some have also questioned whether young people are able to use their vote in a sensible way, asserting that they may lack the necessary maturity (Chan and Clayton, 2006). Both of these issues, turnout and voting among enfranchised 16-and 17-year-olds, can be studied empirically in countries that have in fact lowered the voting age to 16. There is also some debate about the wider impact of civic engagement and how voting at 16 may affect the political debate and the system more widely. Important questions in that regard are what effect a lower voting age may have on young people's attitudes towards democracy and political trust (Tonge and Mycock, 2010). We look at each of these issues in turn below.

Young people's political behaviour
In his influential 2004 book, Voter Turnout and the Dynamics of Electoral Competition in Established Democracies Since 1945, Mark Franklin argues that the decline in turnout in established democracies since the 1960s can be traced back to when the voting age was lowered to 18. Most democracies in Europe and elsewhere changed the age of eligibility from 20 or 21 to 18 in the 1960s or 1970s. Granting voting rights at the age of 18 became the norm in practically the entire democratic world in this period. In terms of turnout, this was a mistake, according to Franklin (2004). By granting voting rights to young people at a time when, in most countries, they are in a transitional phase in their lives after high school, dampens turnout among first time voters. When people do not take part in their first eligible election, they may learn the habit of not voting, which could then lead to a lifetime of abstention or just sporadic voter participation. This is exactly what has happened since the 1960s, according to Franklin (2004), as turnout has declined in almost all established democracies.
Based on this, he hypothesises that a further reduction of the voting age could actually have a beneficial effect on future trends in turnout. Giving young people a chance to take part in democracy at a time when they are still, for the most part, in high school and living in a community that they know could spur higher rates of turnout among first time voters. Young people may then learn the habit of voting, rather than abstention, which over time will lead to rising aggregate turnout levels.
A number of studies have shown that the first part of Franklin's (2004) argument holds water; that 16-and 17-year-olds have higher rates of turnout as first time voters, when given the chance to vote, than 18-and 19-year-olds (Bergh, 2013;Zeglovits and Aichholzer, 2014;Aichholzer and Kritzinger, 2020;Huebner and Eichhorn, 2020;Ødegård et al., 2020). In a new publication, Franklin (2020) takes up the second part of the argument that a lower voting age will over time, due to generational replacement, lead to higher aggregate rates of turnout. By studying trends in voter turnout in countries that have had a voting age of 16 for some time, specifically Argentina, Austria, Brazil, Ecuador and Nicaragua, he is able to conduct the first empirical test of that proposition. The results of analyses of both aggregate and individual level data indicate that indeed there is a substantive positive effect on turnout in these countries in the long run. These effects are somewhat imprecisely measured, because of limitations in the data; the main limitation being that the data cover a relatively short time span in just a handful of countries. We may not yet know the full effect of lowering the voting age on the still young generation of voters who had the chance to vote when they were 16 or 17. However, no negative consequences for turnout from lowering the voting age to 18 were found at these more current reductions to 16; rather there were statistically significant positive effects.
The other aspects of young people's political behaviour that has garnered some interest are their voting or political preferences. Will the granting of voting rights to additional young citizens have a political effect? On this question, the evidence is quite mixed from one country to the next, and even within countries. Franklin (2020) finds that there is a moderate rise in voter volatility when the voting age is lowered. Young people may switch their vote more often than the older voters. In most countries, the young people tend to support centre-left or green parties in somewhat higher numbers than adults, but this is by no means an iron law and support for centre-right and right-wing parties amongst young voters in the last Austrian federal elections was high (ORF, 2019). Also, there is further nuance: in several German states, 16-17-year-olds were often less likely than the general public to vote for parties whose primary position is left or centre-left, but instead often opted for the Green Party at much higher rates than the overall population (Leininger and Faas, 2020). Also, views can indeed change more within this age group, it appears, as Franklin suggested. In Scotland, for example, 16-17year-olds initially were less supportive of Scottish independence than the overall public (Eichhorn et al., 2014). By the time of the independence referendum of 2014, however, many had changed their views with the majority of them embracing independence (Fraser, 2015) at greater rates than Scots overall.

Young people's political attitudes
All types of proposed or implemented changes in electoral law or in the institution of elections, raises the issue of democratic legitimacy. Will the changes have an effect on the legitimacy of elections and on support for democracy, more widely? This is also an issue that has come up in the voting age debate. Sanhueza Petrarca (2020) study the effect of lowering the voting age on political trust and support for democracy in Latin America. She finds consistently positive effects. Voters that were given the right to vote at 16 show higher levels of political trust and greater support for democracy than other voters. Similarly, Aichholzer and Kritzinger (2020) find that Austrian voters at the age of 16 and 17 show greater levels of support for democracy and external political efficacy than other voters. Furthermore, an increase in the overall political interest in the age group could also be observed (Zeglovits and Zandonella, 2013). These findings are suggestive of a genuine benefit of lowering the voting age to 16, however, nuanced differences can be observed here as well.
Research in Scotland, for example, has found that 16-and 17-year-olds in Scotland after enfranchisement had greater levels of several pro-civic attitudes compared with their unfranchised counterparts of the same age in the rest of the UK. But the strength of attitudes was not consistent across all domains. It was more pronounced for perceptions of self-efficacy (especially the question whether 16-and 17-year-olds should be allowed to vote), but less extensive for more general attitudes, such as whether it matters who gets elected (Eichhorn, 2018a). Similarly, insights from experimental Norwegian studies suggest a nuanced picture, too. Some young people in a selection of municipalities were allowed to vote at 16 and could be compared to young people in municipalities where the voting age was maintained at 18. While some positive effects, similar to those found in other countries, were observable (Ødegård et al., 2020), efficacy and political interest in general were not seen to rise in all investigations (Bergh, 2013).
Context may matter, of course. The studies able to identify significant and lasting changes in attitudes were based in countries where we have seen a lowering of the voting age for all elections (such as Austria and several Latin American countries). The countries with positive, but limited effects are cases with partial enfranchisement (such as Scotland where 16-17-year-olds can vote in Scottish local and national elections, while 18 remains the voting age for UK-wide elections) or where young people's enfranchisement was only part of an experiment, rather than a full rollout (as in Norway).
Similar to our discussion about enfranchisement processes, context appears to matter. The sample size is too small to make definitive statements about the precise structure of contextual effects, but it is worth carefully considering that the reduction of the voting age is part of broader systemic questions that may influence how it is experienced and in turn how the earlier involvement of young people may affect broader political debate.

Effects on political debate and the system more widely
Debates about changes to the electoral franchise, even when focussed on just one category, such as age, intersect with wider questions about who should be allowed to participate in deciding about the political structures in a country. Therefore, it should be understood in a contextualised manner (Mycock et al., 2020). Indeed, some arguments against the lowering of the voting age have often focussed on questions about society more widely and in particular the lack of support for such a change among the general population (Electoral Commission, 2003b). In the UK, for example, support for lowering the voting age has traditionally been low with only around one third of the public approving of it throughout (Nelson, 2012), although more recently the opposition appears to have softened somewhat overall (Greenwood, 2018).
However, similar to the question of whether enfranchisement can have an effect on young people at ages 16 and 17, it is also worth examining whether views might change in the overall population after experiencing young people's political engagement. Indeed, the Scottish case demonstrates rather dramatic changes in public opinion. Shortly after the younger people had been allowed to vote for the first time, support for allowing votes at 16 at all elections in the UK rose to 50% in Scotland (Kenealy et al., 2017, pp. 45-76) and even increased further to around 60% (Electoral Commission, 2014). Views of political actors have also evolved. Initially opposed, the Scottish Conservative Party changed their position and voted in favour of lowering the voting age for all Scottish elections in 2015making the decision on the topic unanimous in the Scottish Parliament. In the USA, campaigns for lowering turnout with young people strongly visible in politically engaged leadership roles also tended to increase wider public receptiveness of the idea (Douglas, 2020).
One mechanism through which public perceptions may be affected directly includes the interaction between young people, their families and friends. Research has shown that political socialisation is not a one-way street and that young people indeed also influence their parents (Zaff et al., 2010). This can be enhanced through civic education (McDevitt and Chaffee, 2000). Indeed, there is an indication that this occurs significantly for those at 16 and 17 when being allowed to vote, but the likelihood of young people affecting their families is greater when they also have civic education concurrently in which political issues are discussed in the classroom (Eichhorn, 2018b). How important the intersection between civic education and political attitudes and behaviour is for young people has been demonstrated generally in many studies (Dassonneville et al., 2012). In relation to voting at 16, positive effects discussed above are more likely to materialise when civic education is extensive, as could be seen in Scotland (Kenealy et al., 2017) and Austria (Schwarzer and Zeglovits, 2013;Zeglovits and Zandonella, 2013). However, political education is often a contested arena and not easy to navigate for teachers, if they do not feel confident in being able to moderate discussions of political topics. This is especially relevant in societies where divisions in political views are also reflected in different views amongst groups of teachers, such as in Estonia (Toots and Idnurm, 2020). Given the variety of empirical studies, the difficulty of comparing educational systems and the importance of broader political contexts for the structure of civic education, comparing the precise mechanisms intersecting with the lowering the voting age across countries is very difficult. As Milner (2020) points out, while we can see important links, the precise interplay is not fully understood yet and requires further engagement.

Conclusion
Compared with a decade ago, we are in a very fortunate position. As more countries have lowered the voting age to 16, we have been able to collect empirical data enabling us to study what really happens when 16-and 17-year-olds are allowed to vote in terms of their political behaviour, their attitudes and broader views on their engagement. The richness of data collected in several countries provides us with the opportunity to verify or reject many of the findings from speculative analyses conducted previously.
In none of the countries, for which data are now available, researchers could find negative effects of the lowering of the voting age on young people's engagement or civic attitudes. In many instances the opposite was the case. Enfranchised 16-and 17-year-olds were often more interested in politics, more likely to vote and demonstrated other pro-civic attitudes (such as institutional trust). In many instances, young people enfranchised earlier were more engaged than those classically enfranchised at 18 and longer-term research from Austria and Latin American countries suggests that the effect may at least partially be retained throughout further years of life, resulting in turnout increases. Furthermore, where we have data on public views on the topic, we see support for votes at 16 increase significantly. This may be due to the experience of seeing young people engaged or the influence young people may have on their parents, especially when having had civic education that involved discussions about politics. It seems that the process leading to the introduction of lower enfranchisement ages may play a role as well, seeing popular support increase when more bottom-up approaches are used and young people become visible in campaigns. However, many of the countries that saw earlier enfranchisement in recent years had the process initiated more top-down initially-but in some cases civil society organisations then joined the process of advocating for change.
The research across countries reminds us that lowering the voting age does not happen in isolation of other political and social processes. Indeed, it is important that its relevance is understood in terms of wider discussions of constitutional change (such as in Brazil or the UK). This points also to important differences that we could observe and gaps that continue to exist even after considering the available evidence cross-nationally. The depth of effects on young people's attitudes was not always equal. In some instances, attitude shifts were lasting and wide-ranging (such as in Austria or Latin American countries), but in others attitude shifts may not have gone as far (such as in the case of the Norwegian voting age experiments). We cannot say why those differences in results emerge. It could have to do with the difference in levels of voting age reductions being comprehensive in the former cases from the local to the national level, but only partial for the latter countries. However, to examine this and further questions that try to establish why differences between countries exist, we would need additional and explicitly comparative research.
So far, all the studies cited refer to work that was designed and undertaken within a given national context. This has the advantage that it makes the best use of country-specific data and embeds the work within relevant national discussions (such as those about political education in Estonia). However, it reduces our ability to compare the findings beyond the description of overall dynamics. In particular, it is difficult to study the interplay between different processes that intersect with enfranchisement experiences, such as socialisation and civic education, because the operationalisation of measures used differs greatly between countries. A study that was designed to apply the same methodology across multiple country contexts could provide us with systematic insights currently impossible to gain. This would include the comparison of longer-term effects now that we have enough countries with at least a second round of elections since the first time of enfranchising younger voters. A similar rationale also applies to the study of the processes leading to the lowering of the voting age. While we can compare the origins and processes of the legislative changes made, to properly understand political and civil society motivations for engagement, we would need comparative qualitative work that would allow us to ask similar questions to involved actors across countries.
Work over the past decade has provided relevant insights about what happens when 16-and 17-year-olds are allowed to vote. Overall, the picture appears to be rather positive and bringing together research from different countries has enabled us to identify some overall patterns. However, to understand more deeply what structural factors may affect how precisely earlier enfranchisement interacts with politicalbehaviour and attitudes will require further work of the comparative nature outlined above. That work would enable us to gain insights relevant beyond the topic of lowering the voting age itself and useful to scholars interested in youth political engagement, socialisation and education more widely.