With the Danish general election only a few days away, what can forecasting models tell us about the outcome? Drawing on a model that features economic growth, Richard Nadeau, Michael S. Lewis-Beck and Karina Kosiara-Pedersen predict a comfortable win for a left coalition led by the ruling Social Democrats.
Prime Minister Mette Frederiksen, the centre-left leader of the Danish government, has called for a general election to be held on 1 November. An allied party, the Social Liberals, had threatened a no confidence vote if an election was not held, in light of a critical report on the government’s order for the culling of all mink on health grounds during the Covid-19 pandemic.
Frederiksen hopes to put together a broad coalition of parties, reaching across the left-right divide, thus replacing the one-party minority government of the Social Democrats that she currently leads. This unusual bipartisan gesture comes during troubled times in the country, including roaring gas prices and claims that Frederiksen and her one-party government has been too powerful and opaque, not only when dealing with Covid-19 and mink culling. In the face of these difficult times, some of the polls suggest she should be running scared.
Besides the polls, what does other systematic evidence suggest? To forecast the last Danish (2019) election, we developed in advance a political economy model that successfully predicted the results of that contest. We apply that same model now, updated to forecast the 2022 election. The model bases itself on fundamental economic and political forces acting on the electorate. With respect to the economy, the focus is on economic growth in the period before the election. With respect to political issues, their currents are captured in retrospective aggregate vote intentions. The theory receives testing against data from the past 21 general elections, 1964-2019.
Given the complex multiparty nature of the Danish system, the election outcome to be predicted might not seem obvious. However, the system itself offers some useful regularities. Specifically, there are key Danish parties, accounting for almost 70 percent of the total vote, which have endured over a very long span of time: the Social Democrats and Social Liberals (which can be grouped into a left coalition share); and the Liberals and Conservatives (which can be grouped into a right coalition share).
For the 2022 election, we can measure the size of these two coalitions by simply using opinion polls measuring vote intentions. We employ our standard three-month lead time, with an estimate from late July/early August (averaged from two Epinon and Voxmeter surveys). Based on these surveys, we have computed the expected vote intentions for each of these four major parties as 23.6% for the Social Democrats and 6.5% for the Social Liberals (30.1% for the left coalition), with 14.7% for the Liberals and 12% for the Conservatives (26.7% for the right coalition).
We can see that the Social Democrats would still dominate the left-coalition and that the left coalition would be larger, with a lead of 3.4 percentage points. However, when the left coalition share is calculated as a proportion of the votes for the two coalitions, it comes out at only 53%. This margin, while positive for the left, is narrow and suggests that the race is, indeed, tight, according to the polls.
However, the picture changes when the economy is considered. Elsewhere, the impact of the economy on the Danish voter has been well-demonstrated. The economic measure we focus on here is the economic growth before the election, specifically the three quarters prior (GDP t-9, the annual growth rate three quarters before the election). That growth number shows robustness, at 6.6 percentage points, well reflecting the recovery of the Danish economy from the Covid-19 pandemic.
We ‘synthesise’ these two independent variable measures together in a regression equation, to forecast the left-vote proportion as a function of these political and economic fundamentals. Estimating this equation with ordinary least squares yields a forecast of 58.4% for the left coalition. This strongly suggests a victory for the left-coalition. Why? Because looking at the entire election series, when a left coalition has won, their vote represents, on average, 55.4% of the total support for these four enduring governmental parties. Our current estimate, of 58.4%, is a full 3 percentage points higher than that baseline.
Thus, once economic performance is taken into account, as it should be, a Social Democrat-led coalition appears to be headed for a clear victory. This is, of course, in contradiction to the picture the polls have often painted of a race that is too close to call. That does not seem to be the case now. The bottom line leads to a left victory, with the economy as a crucial factor.
Note: This article gives the views of the authors, not the position of EUROPP – European Politics and Policy or the London School of Economics. Featured image credit: European Union