Cleaning Up an Over-engineered Mess? German Electoral System Reform
Senior Fellow; Director, Society, Culture & Politics Program
Dr. Eric Langenbacher is a Senior Fellow and Director of the Society, Culture & Politics Program at AICGS.
Dr. Langenbacher studied in Canada before completing his PhD in Georgetown University’s Government Department in 2002. His research interests include collective memory, political culture, and electoral politics in Germany and Europe. Recent publications include the edited volumes Twilight of the Merkel Era: Power and Politics in Germany after the 2017 Bundestag Election (2019), The Merkel Republic: The 2013 Bundestag Election and its Consequences (2015), Dynamics of Memory and Identity in Contemporary Europe (co-edited with Ruth Wittlinger and Bill Niven, 2013), Power and the Past: Collective Memory and International Relations (co-edited with Yossi Shain, 2010), and From the Bonn to the Berlin Republic: Germany at the Twentieth Anniversary of Unification (co-edited with Jeffrey J. Anderson, 2010). With David Conradt, he is also the author of The German Polity, 10th and 11th edition (2013, 2017).
Dr. Langenbacher remains affiliated with Georgetown University as Teaching Professor and Director of the Honors Program in the Department of Government. He has also taught at George Washington University, Washington College, The University of Navarre, and the Universidad Nacional de General San Martin in Buenos Aires, Argentina, and has given talks across the world. He was selected Faculty Member of the Year by the School of Foreign Service in 2009 and was awarded a Fulbright grant in 1999-2000 and the Hopper Memorial Fellowship at Georgetown in 2000-2001. Since 2005, he has also been Managing Editor of German Politics and Society, which is housed in Georgetown’s BMW Center for German and European Studies. Dr. Langenbacher has also planned and run dozens of short programs for groups from abroad, as well as for the U.S. Departments of State and Defense on a variety of topics pertaining to American and comparative politics, business, culture, and public policy.
Germany’s electoral system is back in the news because the governing coalition has been considering making some alterations to address consequences of the last set of changes that went into effect for the 2013 election. Even though many people think that such changes are extraordinarily arcane with no real effect, in actual fact, electoral systems are among the most important components of any democratic political system. They perform the crucial function of translating votes from an election into seats in a legislature, and thus directly form the actual balance of party power and governments.
There are vast differences across the world in how this translation process actually works in practice, largely a consequence of the vision of representation countries have chosen. (And this is always ultimately a choice even if to retain a system that was “inherited” from founders or colonial powers). On the one hand, there is the principal-agent vision in which elected representatives are the agents of and have a quasi-contractual obligation to the voters from a territorial constituency. On the other hand, the microcosm of society principle asserts that a legislature should closely resemble the society it represents. The first vision is typically found in first-past-the-post or single-member plurality (also some majority systems like France) such as the UK, India, or the U.S., in which a single winner is the candidate that receives the most votes from territorially demarcated constituencies. The second leads to forms of proportional representation (PR), used in dozens of countries like the Netherlands, Chile, and South Africa.
The first type of system has the advantage of personalized, territorially-based representation, two-party systems, single-party governments, and a simplicity both in voting (x marks the spot) and in tabulating results. But, they have high levels of disproportionality (meaning lack of vote-seat correspondence). Often, the “winning” party receives 10 to 20 percent more seats than its vote share and governs with such a “manufactured” majority despite not winning a majority of the vote. For instance, in 2019 Boris Johnson’s Conservative Party won 56.1 percent of the seats in the UK House of Commons based on 43.6 percent of the vote, a 12.5 percent boost. By contrast, proportional representation produces multiparty systems with a high degree of vote-seat correspondence. But, they can suffer from a fragmented, unstable party system, and a lack of politicians responsible for a given territorial unit such as a city or county. As discussed below, their greater complexity can present some comprehensibility challenges to voters and delays determining winners and eventual governments. Indeed, coalitions are common and can take months to form—over five months in Germany after the September 2017 Bundestag election.
Parties are concerned about the overall functioning and legitimacy of the political system, but they are also self-interested.
Parties understand the power and consequences of electoral systems. Of course, parties are concerned about the overall functioning and legitimacy of the political system, but they are also self-interested. They will advocate for the status quo or changes depending on their own utility. The Canadian Liberals supported the introduction of PR, but then dropped the idea quickly after they benefitted from the current system. Unsurprisingly, many in the British Labour Party are currently advocating for PR. Larger parties often prefer plurality systems because they benefit, squeezing out smaller parties (Duverger’s Laws). Both the SPD and CDU flirted with moving to such a system in the 1960s to no avail. Smaller parties advocate for a more proportional system—and the smallest would like lower or no electoral thresholds. The reality of partisan self-interest is also a factor in German debates about the electoral system.
The founders of the Federal Republic pioneered an electoral system in 1949 that combines both visions of representation. Called a mixed system or the personalized proportional system (personalisierte Verhältniswahl), a proportion of the seats are elected through single-member plurality (direct mandates) and another portion is elected through proportional representation or PR (list mandates). Half of the legal minimum Bundestag seats (currently 299) are elected from small, territorially-compact constituencies in which the winner is decided through plurality with the so-called first vote (Erststimme). The drawing of constituency borders is tightly regulated and gerrymandering is not considered a problem. The other half of the seats is allocated according to closed party list (the party decides on the candidates and the order) proportional representation (using the Sainte-Laguë method) with national level vote totals (Zweitstimme), but from within each of the sixteen Bundesländer. Each state gets a number of constituencies (Wahlkreise) proportional to its population and then exactly the same number of list PR seats: Bremen has two constituencies and two PR seats for a total of four; North Rhine-Westphalia (NRW) has 64 constituencies and PR seats for 128 total. The second vote is the more important of the two and determines the actual power balance in the Bundestag.
To simplify the seat allocation process, essentially the authorities calculate how many seats each party would get in each Land as if all seats (in NRW, for example, 128) were allocated based on the national party vote totals. Then, they subtract the direct mandates each party has already won from this theoretical total to get the actual seats the parties get from their list. Usually, the biggest parties win most or all the direct mandates, so smaller parties do very well with the second-vote seats. In the end, the overall result is quite proportional, which is why some analysts refer to the system as a PR one and not “mixed.” Others call the German system “mixed member proportional,” contrasted with “mixed member majoritarian” or parallel voting where the two portions of the election are separately determined and then added together.
There are several complications. Since 1953, there is a national 5 percent threshold (Sperrklausel). Parties need to surpass this to be eligible for seats in the Bundestag—this most recently impacted the FDP and AfD in the 2013 election when they came in just below the threshold and received no seats. In 1949 this applied within each state and in 1990 there were separate thresholds in eastern and western Germany (benefitting the Greens). If a party receives three direct mandates, they are eligible to get PR seats based in their share of the vote (last benefitting the PDS in 1994). Also, any direct mandates won are kept and the party is seated in the parliament (even if it failed to reach the threshold—as with the PDS’s two seats in 2002). Finally, there are so-called overhanging mandates (Überhangmandate). Sometimes a party will win more direct mandates than they would receive in a perfectly proportional allocation process. When this happens, the size of the Bundestag is increased. For instance, in 2002—the first year of a reduced minimum size (598 from 656)—there were five such mandates and a resultant Bundestag with 603 seats.
This system worked extraordinarily well over the decades, delivering a moderate multi-party system with high levels of proportionality. In fact, the system was so stellar that it inspired similar reforms in scores of other countries, such as New Zealand, Bolivia, Russia, Japan, and Mexico. But, in recent cycles, problems started to mount—especially with overhanging mandates. For most of the postwar period, they were uncommon. In fact, there were zero between 1965 and 1980. One reason why is that the party system was relatively uncomplicated with the two catch-all parties (known as the elephants) CDU/CSU and SPD dominating the second vote (together receiving 91.2 percent of the second vote in 1976, for instance) alternating in power in coalition with the FDP. Even after the Greens entered the Bundestag, these patterns were not significantly disrupted.
As the graph below shows, overhanging mandates began to visibly increase after reunification and accelerate around the turn of the century—a consequence of more parties gaining representation and the concomitant fall in the vote share of the two elephants, sinking to 53.4 percent in 2017. When the first and second voter totals were more aligned, overhanging mandates were rare. Most generally, over time, the competing logic of the two partial electoral systems—two parties dominating the direct mandates and rising multi-party dynamics from the PR component—started to create real tensions, disparities, and more disproportionality (less voter-seat correspondence).
Table 1: Overhanging and Compensatory Mandates in the Federal Republic
Given these developments, the Federal Constitutional Court (Bundesverfassungsgericht) intervened in 2008 and 2011, mandating changes to the electoral law. It found that the original system was benefitting larger parties (creating disproportionality) and there was also the danger of a “negative voting weight” (negatives Stimmgewicht). This is rather complicated, but, essentially with the seat allocation process taking place at the state level, in some scenarios with overhanging mandates more second votes for a party can lead to fewer overall seats for that party. As a consequence, the Bundestag revised the system in time for the 2013 election, adding so-called compensatory seats (Ausgleichsmandate) so that there is a closer vote-seat correspondence and so that the national vote-seat ratio is the same (each voter has equal weight). Already rather complicated before, the Bundestag responded with yet more complexities. Few regular voters find the new system to be anywhere near intelligible.
Nevertheless, this change did not substantially affect the size of the Bundestag in 2013—there were 33 overhanging and compensatory mandates. But, in 2017, the number skyrocketed to 111 and a Bundestag with 709 seats. This makes the Bundestag the second-largest legislature in the world, behind only China with about 3,000 members. Even the European Parliament, representing 446 million people presently, has only 705 members. The extra deputies are costing taxpayers approximately €50 million more per year. In 2020, almost €1 billion (yes, billion) of taxpayers’ money will be spent. They had to redesign the plenary chamber to accommodate so many members—and had to find more office space, hire staff, and pay for salaries and benefits.
Overly large chambers can make it difficult to negotiate, can lead to tension because so many cleavages and issues have found specific representatives, and will likely produce more and unnecessary regulations.
Moreover, political scientists have developed a rule of thumb about the ideal size of lower houses—the cube root of the population. Overly large chambers can make it difficult to negotiate, can lead to tension because so many cleavages and issues have found specific representatives, and will likely produce more and unnecessary regulations. (Conversely, overly small chambers can fail to represent all the shades of political opinion, produce a higher voter to representative ratio, making it too difficult and onerous for voters to communicate with representatives). By this measure, Germany has long had an excessively large legislature. According to this rule, it should have a chamber with only 435 members. By the way, some projections see a Bundestag with over 800 members after the 2021 election under the current rules!
For all of these reasons, there has been recent discussion about yet another set of changes to the electoral system to reduce the number of members, but also to maintain proportionality (and thus constitutionality). Parties, as always, are also pursuing their self-interest. Several opposition parties (Greens, Left, and FDP) introduced a proposal in late 2019 to reduce the number of constituencies to 250 and increase the number of list seats to 380 (about 60 percent of the total), obviating the need for compensatory mandates. The coalition parties slow-walked this idea—although the SPD appears more open to reform. Especially the CSU, which wins almost all direct mandates in Bavaria, is against reducing the number of constituencies—also arguing that the resulting larger populations will make it more challenging for deputies to deliver constituent services. The coalition did have another proposal—reducing districts by nineteen and allowing seven overhanging mandates not to be compensated—but this petered out just before the summer break in early July 2020. Although talks may continue over the summer, there is little chance that anything will change in time for the 2021 election. Parties are already involved in recruitment and formulating their state lists.
The system’s performance has declined since reunification and change was mandated by the Constitutional Court.
The mixed German electoral system has long been exemplary—combining the best of the two visions of representation. It has rightfully served as a global model and I sincerely hope that the basic two-vote structure is retained. Nevertheless, the system’s performance has declined since reunification and change was mandated by the Constitutional Court. Whatever the cause—party self-interest, competing proposals, inertia, etc.—the medicine has been worse than the disease. The current system is an over-engineered mess and the immediate prospects to clean it up are dim. Even more worrisome is the prospect that the Court will eventually invalidate the 5 percent threshold (as it has for elections to the European Parliament), creating an even more fragmented parliament and necessitating additional legislative responses. In any case, we will be talking about the German electoral system for quite some time—at least until the 2025 election.
 Order matters—if a party will get, say, 10 seats from the second vote allocation process, the top 10 names will sit in the Bundestag. Parties can engineer their lists to ensure that top politicians are highly placed—as well as ensuring that women (and potentially) minorities have higher places and thus seats in parliament.