However, analysis of the Audit of Political Engagement, drawing on fifteen years of data spanning the period of New Labour to the present day, reveals that the winner-loser dynamic has not just weakened, it has actively reversed.

Since the beginning of the Audit, we have asked people – in a question unique to Audit surveys – whether "When people like me get involved in politics, they really can change the way that the UK is run". Consistently, barely more than a third of the public agree with the statement.

But new analysis of the data shows a marked change in the response of supporters of the governing and opposition parties to this question since 2004. People who support the government of the day are increasingly less confident that they can make a difference, while people who support other parties, and even those who would not vote at all, have higher efficacy than ever.

In the graph below, supporters of the governing party (or parties, in the case of the Conservative-Liberal Democrat Coalition in power at the time of the 2010-14 surveys), are less and less likely to express feelings of efficacy. Crucially, this happens within the period that a party has government status, as well as between switches in party control.

Meanwhile - beginning, it seems, around 2008 – those who did not support the governing party (or parties) tended to express higher and higher levels of political efficacy. By 2015, government supporters were no more likely to feel that their involvement in politics would be effective than the rest of the population; by 2006, they were actually less likely.

Secondly, Labour supporters appear to feel more politically efficacious in opposition since 2010 than Conservative supporters did under the New Labour governments. In the Audits published in the 2004-2010 period (based on November/December polling in the years 2003-2009), net agreement among Conservative supporters lagged Labour supporters by an average of 19 percentage points. However, since the 2010 general election, Labour supporters in opposition report, on average, a much higher sense of political efficacy, not just compared to how Conservative supporters felt in opposition, but also how Conservative supporters have felt since their party returned to government.

Insofar as people are feeling more politically efficacious, the trend among opposition supporters can be seen as a positive development. However, the ‘improvement’ we observe may be confined to people who are already inclined to participate in politics. To test this, I separated out the specific trend among non-voters. The results are striking: non-voters also report a higher sense of political efficacy, in arguably a more dramatic way. In the 2000s, net agreement with the question about the eficacy of getting involved in politics was an average of -41%; in the 2010-2016 period, this has improved to -24%. It is an improvement nearly twice as large as that which occurred among non-governing party voters. However, non-voters are still somewhat more likely to disagree that they can effect political change, which is likely to form a barrier to their participation.

The data illustrates that a change has taken place but cannot fully explain why. It may be that, in recent years, the ‘winners’ and ‘losers’ are not as clear-cut as they were at the start of the Audit research.

Labour governments prior to 2010 were much more clearly in the saddle compared to the Conservative and Conservative-led governments that have held office since. The latter administrations have often been forced into political U-turns – whether due to upset among their own backbenchers, coalition partners, or strong mobilisation by the opposition.

In the coalition years, the ‘constraints and tensions of being in coalition’ wore on Conservative supporters; as the British Social Attitudes survey showed in 2015, by the end they were much less supportive of the idea of coalitions than they were in 2010 (NatCen, 2015). The dynamic of coalitions has to some extent continued in the subsequent governments, which have been unable to push through ‘true blue’ reforms which might excite the party’s base. While the Blair and Brown governments were not immune to big policy u-turns of their own, very different parliamentary arithmetic has changed the extent to which ‘winners’ really win and ‘losers’ really lose.

It may also be that new platforms for political engagement such as e-petitioning give supporters of opposition parties or no party at all a vehicle for political influence hitherto unavailable to them, even if this is often perceived rather than actual influence. As the Audit of Political Engagement shows, after voting e-petitioning is generally the most popular form of political participation that people say they have done and would be willing to do in the future.

The implications for contemporary politics are not entirely clear. However, as the academics Gastil and Xenos, 2010 have shown, there is a well-documented causal relationship between efficacy and participation of various kinds, such as volunteering for parties and single-issue groups, or attending political events. If the trend in the perceived political efficacy of opposition party supporters were to grow still further, then it is possible we might see a steadily larger group upon whose time, effort, money and support parties and campaigns could potentially draw. Finally, given that even those who don’t vote are today somewhat less dismissive about the effects of getting involved than they were at the start of the Audit study, there may be more potential for the activation of non-voters into patterns of political engagement than expected.