Will the Labour party turn into a losing machine under Corbyn?

Jeremy Corbyn is currently the frontrunner for the leadership of the British Labour Party. While few would have bet on him when he announced his bid, he is now the bookies’ favourite.

Labour leadership candidates by Google searches, May-August 2015
Google searches for Labour leadership candidates, May-August 2015 (Google trends)

A number of prominent figure on the right of the party have come forward to warn against the electoral consequences of him becoming leader. Tony Blair wrote that Labour would face “annihilation” if the left-winger were to lead the party, and Gordon Brown explained at great length how a Corbyn leadership would affect the party’s credibility. The general theory among pundits is that only a centrist candidate can be able to win elections in Britain. There are three reasons behind this. The first is institutional: in a majoritarian system, you win elections in the centre, and you need to convince the voter in the middle of the voter distribution. The second and third are historical: Tony Blair managed to win three consecutive elections on a centrist (or centre-right) platform; and Michael Foot faced a mortifying defeat in 1983 with a clear left-wing platform. But how does the electoral score relate to the left-right position of parties?

In the Graph 2 below (click for interactive version), I have plotted the vote share of the Labour and Conservative parties between 1945 and 2010 against their left-right position as measured by the comparative manifesto project. In other posts I have used the Chapel Hill Expert Survey, but the manifesto data goes back much further in time. Regression is faced with a number of problems: endogeneity (it is difficult to know whether the ideological position determines success, or whether electoral success in previous elections conditions the electoral position), measurement problems (the British party system has become more multipartisan over the last 30 years, so vote shares have declined for both parties). The correlation is very weak, and the relationship is not significant. The regression line (you can see it if you click on the graph) slopes downwards for both parties, but it has probably more to do with the fact that both parties were more left-wing in the 1950s, and also gathered more votes, than being more left-wing yielding more votes.

Left-right position and vote share of British Political Parties, 1945-2010

However, the representation of the data gives an interesting representation of the history of British politics since 1945. In the postwar period, both major parties were on the left according to the data. That’s the period in which the NHS was created and major industries were nationalised. Throughout the 1950s and 1960s, the difference between the two parties was not that big (Graph 3). This changed to a great extent in the 1970s, when the distance between Labour and Tories increased and a major period of polarisation started. The Thatcher era is of course the most symptomatic of this: the Tories veered to the right, while Labour engaged in soul-searching with large swings up and down (Graph 3). The Blair leadership shifted the party to the right, bringing back electoral success. The distance between the two main parties decreased again, but this time, convergence took place on the right rather than on the left, as in the 1950s and 1960s.

In graph 2, it is striking to see two parallel lines going right and upwards: the Tories between 1974 and 1979 (Thatcher) and Labour between 1992 and 1997 (Blair). In both cases, a major shift to the right coincided with a major increase in vote share (following crushing defeats). In contrast, the two instances where the Labour party veered to the left (in 1974 and 1983) coincided with lower vote shares. A move to the left did coincide with a higher vote share between 1987 and 1992, but this was not enough to secure a majority in parliament.

Left-right position of Conservative and Labour party, 1945-2010 (Comparative Manifesto Data)
Left-right position of Labour and Conservative parties, 1945-2010

Does better research mean better teaching?

Does research performance increase the quality of teaching, or does it hamper it? Many people say that there is a positive relationship between research performance and teaching quality. Research should feed back into teaching and vice-versa. Following this logic, there should be a clearly positive relationship between student satisfaction (with the obvious problems of taking this as a measure of quality) and research output. However, there are also obvious trade-offs between the time necessary to deliver good teaching, supervise students, etc. and that necessary to do the things that are considered in research rankings: writing articles in high-ranking journals, or preparing research grants.

Since the results of the National Student Survey for 2015 has been released a few days ago, I have tried to see how NSS scores relate to research performance as measured by GPA score for the 2014 Research Excellence framework for Politics departments. The graph below plots UK Politics departments by overall satisfaction scores among undergraduate students in the NSS (question 22 of the NSS) as taken from here, and their GPA score (the average number of stars for research output, impact and “environment”) as taken from here. Here’s how it looks (you can access an interactive version by clicking on the graph):

UK Politics Departments plotted by NSS and REF Scores
If you draw a regression line, it actually slopes downwards, which means that departments with high research scores also have on average lower student satisfaction scores. However, the relationship is not significant and the correlation is weak (R-square-0.05). It may be driven by the London factor: UCL, LSE and King’s have low student satisfaction scores but high GPAs. At the other end, Liverpool Hope and Surrey did extremely well for student satisfaction but didn’t fare very well in the REF. Essex, Oxford and Warwick do well on both counts. As a whole, this casts some doubt on the idea that better research means better teaching, with the caveats linked to measurement in both areas.

An Interactive Map of West European Parties

Following on the previous post on Southern Europe, this is a map of all West European parties mapped along their economic policy positioning and their position on European integration. By clicking here, you get access to a large interactive version of the map where you can show and hide countries and zoom in and out. As for the previous one, data is based on the latest round of the Chapel Hill Expert Survey.
West European Political Parties by Ideological Positioning

A Map of the South European Party System

There has been a lot of discussion recently about the rise of left-wing Euroscepticism in Southern Europe in the midst of the Eurozone crisis, with parties like Syriza and Podemos conquering power (in the case of the former) or appearing as serious challengers to mainstream parties (in the case of the latter). In the graphs below I have used recent data from the last round of the Chapel Hill expert survey 2014 on parties’ ideological placements, combined with the latest poll results I could find as of August 2015. The horizontal axis shows the position of parties on the economic left-right dimension, and the vertical axis their position regarding European integration. The size of the bubbles corresponds to to the electoral strength of each party as measured by polls. Here’s how it looks:

In spite of the assumed similarity of party systems in Southern Europe, we can see some clear differences between them. There is nothing incredibly new here, but it visually represents things in a nice way. Greece and Portugal appear as somewhat opposite poles in terms of their fragmentation and their centre of gravity. The Greek party system is more fragmented and its centre of gravity has shifted to the lower left corner of the quadrant, that is, the Eurosceptic left. In contrast, the two major Portuguese parties are situated towards the upper right corner, or the pro-European right. The two Portuguese parties in the lower left, the Communists and the left bloc, have remained relatively weak compared to their Greek and Spanish counterparts. The Spanish party system is in-between these. Unlike PASOK, the PSOE has remained strong even if Podemos represents a new force on the Eurosceptic left. The Italian party system is clearly different from the others. It contains a number of significant political forces in the Eurosceptic economic right quadrant, a bit more like what you would probably see in Northern Europe.

Human capital and tourist scams

A few days ago I arrived at Lisbon airport, took a cab (registry nr 283) for a journey I’ve done countless times and which usually costs a bit less than 10 Euros. I chat in Portuguese with the cab driver, give him directions, signalling that I know my way around. The meter indicates 9 Euros when we arrive. After offloading luggage, the driver *switches to English* and says it’s 30 Euros, that from the airport it’s a special tariff, pointing at a page of a prospectus of unknown origin (maybe the fares for Oslo airport?). I say – in Portuguese – that I’m paying what’s on the meter, which I do after yelling a few things at him.

What upset me most was not that he was a crook, but that he was so rubbish at it. How is this country gonna get out of the recession if even the tourist scam business lacks the appropriate skills? I am all for conning German and British tourists, but I clearly signaled that I was Portuguese and knew my way around. I was almost tempted to give him advice on how to choose targets, hide the meter in some way, etc. At least do it right, for God’s sake.

How did David Cameron manage to win the 2015 elections in spite of austerity?

The poor hit by austerity don’t vote, while the rich who benefit do

How can we explain the Conservative victory in last week’s British elections in the context of the austerity measures the Tories have been pushing through since they came to office? Indeed, even if the deficit hasn’t come down near the levels announced by the Coalition, there have been net cuts in social spending since it came to power. Political economy theories usually argue that voters sanction governmnets who cut social benefits. Building on the good ratings for economic competence attributed to the Tories in opinion polls, “austerians” have been trumpeting that this is a vindication of the austerity agenda. The storyline is that austerity works for everyone, unemployment is down, and this is why the Conservative party has triumphed in the polls. However, there is clear evidence that not everybody has been better off under the Coalition austerity plans. These plans have had distinct distributional effects: the poorest households have been hit the hardest, while middle-and higher income households have come out relatively unscathed.

This victory is interesting because the common wisdom until recently was that politicians would systematically avoid cuts in public spending that would challenge their electoral prospects. This is especially important for welfare programs since the path-breaking analysis of Paul Pierson: people like social programs they are entitled to, and getting rid of them loses votes for incumbents. However, cuts in public spending didn’t hamper the electoral score of the Tories here, and I think this is due to the setup of the British welfare state and the composition of the Tory electorate.

In a nutshell, low-incomes that paid the price for welfare cuts don’t vote, and especially don’t vote Tory, while those on higher incomes for whom lower taxes are more important do vote. After the poll debacle of the election, it appears that it wasn’t “Shy Tories” that were the decisive factor, but potential Labour supporters not bothering to vote. In the graph below, I have put together two measures: the share of net income coming from the state (from here; negative values mean transfers to the state as a share of net income), and the likelihood to vote in the general elections from the British Election Study by income quintile. The latter is not a very good measure because people systematically lie when they say whether they’re going to vote. However, it reports the difference to the average across quintiles on a scale from from 1 to 5. What the graph shows is that people who receive most of their income from the state via social transfers (and who are the most likely to be hit by austerity cuts) are also those that are the least likely to go and vote. By contrast, those who are net contributors to the public budget – and have an interest in cuts to get lower taxes – are those that are the most likely to vote. As I argued elsewhere, higher income quintiles are also much more likely to vote Tory. Hence, austerity could be pursued while limiting a potential backlash because it harms those that don’t vote, and rewards those who do.

image (66)Besides this structural tendency observed about everywhere (the poor are much less likely to vote than the rich), this is also connected to some specificities of the British welfare state that make it easier for Conservative governments to retrench social programs without facing electoral sanctions. Because social programs are strongly targeted at the poor via means-testing and flat rates, the middle class basically doesn’t have an interest in welfare. The drop between middle class wages and benefits is simply too large for this group to consider it a valuable safety net. Figure 2 shows how jobseekers’ benefit is set at much lower lowers than in most other European countries. In countries of continental Europe, where the goal of social benefits is not only to preserve a basic safety net for the poor but to maintain middle class incomes in periods of unemployment or sickness, it is much more difficult to retrench benefits because the middle class also benefits from them. This is why you see people taking to the streets in France about every time the government seeks to downsize entitlement or benefit levels. In Britain, cutting them doesn’t cause much of a fuss because those that suffer don’t vote, and the middle class prefers lower taxes to social insurance.

Finally, it must be noted that the schemes that benefit the middle class as well, such as pensions and the NHS, were not cut. Pension spending was ring-fenced by the government (who was probably aware of the much higher turnout of older voters), while NHS spending is meant to increase in the parliament, probably at the expense of schemes benefitting groups with low turnout.

image (67)

Welfare States and Migration Policy Regimes in Europe: How More Welfare Reduces Immigration Flows

I gave a talk on the relationship between welfare and immigration at the ISCTE-University of Lisbon last week. The talk can be listened to below and the slides are here.

Does fiscal austerity strengthen UKIP?


In their manifesto published yesterday, the Conservative party have pledged to increase funding for the NHS by 8bn a year, besides other policies that have been criticized for for not being clearly funded. Many observers have argued that all these spending increases will need to be be compensated somehow by deep cuts in other domains, such as the welfare budget, where 12bn are to be cut by 2017-2018. In this respect, the Conservative party will be continuing the policy of reductions in public spending which some believe will lead to levels unseen since the 1930s. Now another thing that the Conservatives want, it’s to prevent the rise of UKIP. Can these two things be reconciled? In the graph above, I have plotted together public spending per capita in 2012-2013 (from here) and UKIP voting intentions from the British Election Study. There is a surprising level of fit (72% of the variance explained), and the clearly negative relationship holds if we control for unemployment (data here: UKIPSpending). The less public spending there is, the more UKIP support there is. Basically, if one were to assume that this relationships was causal, it would “cost” 200 pounds of increased spending per capita to reduce the UKIP vote by 1%. I am not sure whether this fits in the long term economic plan.

The irrelevant British median voter

When you watch the campaign for the next general elections in the United Kingdom, you may get the impression that the two biggest contenders, Ed Milliband and David Cameron are competing for the same voters. They promise this, they promise that in multiple attempts to convince what in political science jargon we call the “median voter”, that is, the person in the very middle of the voter distribution, the one that decides elections. In theory, this is how it works in majoritarian systems. In fact, it’s not.

In the theory, the median voter swings from one side to the other, depending on the promises made by the different parties (there is a movie with Kevin Costner about this). In a perfect bipartisan system, one vote lost for one party is a vote won for the other party. This means that the vote share of one party is more or less perfectly correlated with the votes for the other party: 1% more for one party means 1% less for the other party. This describes very well the relationship between the Democratic and the Republican vote in the US, where there are only two relevant parties. The graph below plots together the vote share of the two main parties in Congressional elections between 1960 and 2008. All points (years) are very close to the line. The R-square statistic means that 82% of the variation in the Democratic vote can be explained by the Republican vote. If the Republican vote increases by 1% of the total vote, the Democratic vote is lower by 1.18%.

GraphUS1Now look at the same graph for elections in the United Kingdom. There is no correlation whatsoever between the level of the Conservative Vote and the level of the Labour vote. The R-squared is close to zero, and the results are not statistically significant. This means that the level of the Labour vote cannot be explained by the level of the Conservative vote: when the Labour vote goes down, the Conservative vote doesn’t go up. They are completely unconnected to each other, which seems to imply that that there are very small voter movements from Labour to the Conservatives (this is confirmed by British election survey data). When the Tories lose votes, they go elsewhere.


Where do these votes go? I have run the same kind of analysis drawing on the UK polling report, which brings together 1913 polls since May 2010. That a very large number of polls which allows to run a few analyses with a higher degree of confidence than election results. What do they show? First, the pattern highlighted above between Labour and Conservative votes is similar: there is only a very weak relationship: only 1% of the variation in the Conservative vote can be explained by variation in the Labour vote. If anything, the relationship is actually positive. If the Labour vote goes up by 1%, the Conservative vote goes up by 0.1%. When Labour loses votes, they don’t go to the Conservatives, and vice-versa. On the opposite, they tend to go up together, even if the association is very weak.



Now look at the relationship between Conservative and UKIP voting intentions. The relationship is much clearer, and the dots are much closer to the line. 20% of the Conservative vote can be explained by the UKIP vote. When the UKIP vote goes up 1%, the Conservatives lose on average 0.26%.

GraphUKIPConsevative4 Something similar happens with Labour: its voting intentions are negatively associated with Greens and UKIP voting intentions. The association between the Green and Labour vote is actually even stronger: R-squared is 0.51, which means that 51% of the variation in the Labour vote can be explained by the Green vote. The relationship goes in a similar direction with UKIP. When the UKIP vote goes up 1%, the Labour vote goes down 0.76% on average, and 58% of the Labour vote can be explained by the UKIP vote.


What about the Libdems? The Libdem vote is positively associated with the Conservative vote (perhaps that’s what happens when you enter a coalition) (R-square: 0.12) but unrelated to the Labour vote.

GraphConservative Libdem


A few caveats: OLS regression as used here do not seem to be a good method to analyse and predict votes in multiparty systems. However, graphically this provides some insights on the dynamic of electoral competition in the United Kingdom: parties compete for votes within rather than across ideological blocs. Few people switch from Conservative to Labour or from Labour to Conservative. More people switch from Libdem to Labour or from Labour to SNP (within the left) or from Conservative to UKIP (within the right). This tends to indicate that contests between the two main parties and their smaller challengers on each side of the political spectrum are more important than between them. Politics is moving to the extremes and is no longer fought in the center, if it ever was.

The Political Economy of the British General Election

The real campaign for the British general election is now in full swing and there’s a new poll every hour predicting a lead of one or the other party. Now polls change all the time, and most of the time they are wrong. I don’t care that much about how many votes Labour gets today as compared to yesterday. What I’m interested in, however, is the broader dynamic of representation across income groups and social classes, which tend to be fairly stable. Last weekend I have been playing with some data from recent waves of the British Election Study and drawn some graphs about the appeal of the different parties across income groups, and the composition of their electorate by income group. A few interesting facts:

1. Income matters a lot more for the Tories and UKIP than for Labourimage (35)

The Graph above (large interactive version here) shows the share of party support among income groups (yearly gross household income) in Wave 3 of the British Election Study. The lines are quite different for the big parties: support for Labour is fairly stable across income groups, and only drops among the rich, from 60’000k onwards. Labour looks like a cross-class party where increases in income do not significantly raise or decrease support, at least until a certain threshold. The picture is very different for the Tories, where there is a much larger difference in support between voters across levels of the income distribution. Poor voters are much less likely to vote Tory, while rich voters are much more likely to do so. In some way, this can explain the policies that the government is pushing through, namely cutting benefits at the bottom (who do not vote Tory anyway) to lower the fiscal burden at the top (who do disproportionately vote Tory). Interestingly UKIP has a line going in the opposite direction: UKIP support declines steadily as income increases. It has the working-class profile that Labour doesn’t (no longer) have. Note the high support for the Greens among voters a the very bottom of the distribution, which may be due to the high proportion of students.

The second graph (below and interactive here) shows the difference in support between the different income groups and the mean. What the graph means is that the difference in Tory support between people with an household income of 150’000 pounds and higher and the whole electorate is 120% (this is a difference; they are 2.2 times more likely to vote Tory than average). Once again, the trend described above is visible: income groups are very polarised in Tory (and UKIP, which is the mirror image) support, while the difference is small for Labour

image (36)

The graphs above don’t tell you how important these different income groups are electorally. The very rich vote to a much larger extent for the Tories, but they may be a small minority. The graph below shows the size of each income group in proportion to the overall electorate. The larger income groups are located between 10’000 and 40’000 pounds a year. In all of these, Labour wins, while the Tories clearly dominate in the higher ones, but they are much smaller.

2. Different compositions in the electorate create different policy incentives

This is not very surprising but it is nice to see in the data. In the graph below I have divided the voters of the main parties in three income groups of similar size in the overall electorate (I have recoded the categories above). It could represent the working class, the middle class and the upper class. When parties devise policies, they seek to forge alliances between these different groups, which is likely to be at the expense of one group (e.g the rich want low taxes and the poor want more spending, but it is difficult to do both). Within Labour, the lower class is the biggest, while the upper and middle class are close and equally sized. Within the Tories, the upper class is by far the biggest, while the lower class is the smallest. Internally, an alliance between middle and upper class has a clear majority within the Conservatives, while this alliance fails to do so within Labour. Instead, an alliance between lower and middle class has a majority (last graph). In a very stylized way, this internal balance of power explains fairly well the policies that the different parties offer to voters.