Does Hungary fulfill the requirements of democracy given by the facts on the new voting system?

Question

We had an election with a new voting system which was created by the governing party. The election system works like this: You can vote for electorates and party list. Half of the parliment comes from electorates, rest comes from list.

And the basic methodology through an example, is:
Party A got 4000 votes
Party B got 3000 votes
Party C got 1000 votes
Party D, E got 200-200 votes.

Party A receives 1000 votes (A-B) for list votes
Party B receives 3000 votes for list votes
Party C receives 1000 votes for list votes
Party D,E receives nothing since they didn't collect 5% globally (or if they are alliances by 2 or 3+ parties, they need at least 10 or 15%).

My concerns are the followings:

• First of all, the current results, which is the fact that the Party A got 45% of votes, but it gets 67% of the seats. Which means the rest (55% of votes) have no chance to unifiy, since they received the 33% of the seats. Here is the result in Hungarian.
• Theoretically this system might result that a single 5% party could take 100% of the seats over 6 alliances with result of 14.99% of votes which will be ignored by the system.

Does this system have any democratic election characteristic or even with these anomalies is it a democratic system?

Answer 1

Many democratic countries seem to have such properties in their election systems.

Split "sources of seats" (electorates/lists) aren't seen often, but they happen and aren't a problem as such, I won't argue if it's good or bad, but it's not a drastic condition. Electoral area systems do lead to a "winner-take-all" scenario, where a 55% majority often leads to a total dominance - a visible example is the USA elections; where if 51% of a state votes for a presidental candidate, then nationally 100% of that state votes go for him/her.

A policy that assigns no seats for parties below a certain threshold is a well-established paradigm - see http://en.wikipedia.org/wiki/Election_threshold, most European countries have such a feature. It does provide an incentive to vote for parties where "your vote matters", i.e., who are likely to get above 5%; and not to vote for very small upstarts, which is a bit restrictive; but it has benefits in getting a functioning government that's able to get at least something done; as before those limits (early 20th century) it was observed that lack of such limits often results in a too fragmented parliament that simply can't agree on any government at all.

Answer 2

Mixed-Member Proportional Representation

Hungary used Mixed-Member Proportional Representation (MMP), which is not an unusual system — Germany uses it for the elections for the Bundestag, as do Scotland, Wales, and New Zealand, among others.

As CGP Grey explains in his excellent series on voting systems:

https://www.youtube.com/watch?v=QT0I-sdoSXU

Hungary

However, the Hungarian system has been tweaked a bit between the 2010 and the 2014 elections. A series of articles in the New York Times documents what happened.

Fidesz started immediately after its election victory in 2010 to reshape the electoral system to ensure its hold on power. The Fidesz parliamentary bloc, which enacted constitutional changes without including or consulting any opposition party, slashed the size of the parliament in half, redrew all of the individual constituencies unilaterally, changed the two-round system to a single first-past-the-post election for individual constituencies, and altered the way votes were aggregated.

_{"Hungary: An Election in Question", New York Times}

Some of the changes were agreed upon by all parties, such as the reduction in size of the parliament from 386 to 199 and the accompanying redrawing of districts.

The most important and controversial change was that of the way "lost votes" are calculated and aggregated. In most MMP systems, if a party loses the vote for a constituency, the votes go towards the list. However, in the Hungarian system, surplus votes of the party winning the constituency go towards the list as well, skewing the result away from proportionality.

Professor Gábor Tóka concludes that while the changes are questionable, they are not unconstitutional.

(...) since the constitution, unlike many of its European counterparts, does not prescribe proportional representation in elections, no feature of the new electoral system discussed so far appears to be unconstitutional.

_{Constitutional Principles and Electoral Democracy in Hungary by Gábor Tóka}

However, he estimates that

(...) a united center-left opposition might get 8% fewer parliamentary seats than Fidesz if both got an equal share of the votes.

How democratical that is, is left for the reader to decide.

Answer 3

I really didn’t understand the voting system from your brief description, so I checked with a (German language) source that is typically very trustworthy when it comes to the intricacies of voting systems (Wahlrecht.de) and found that … your description is somewhat incomplete.

So first, how does it work? (Paraphrasing from the Wahlrecht.de description)

1. A voter has two votes: one for a local candidate elected by a simple majority and one for a national list. (Voters not residing in Hungary only receive a national list vote.)

2. As expected, local candidates are elected by a simple majority; this fills 106 of the 199 seats in parliament.

3. The remaining 93 seats are distributed according to the national list vote shares using the d’Hondt method (also known as Jefferson’s method). However, not only the national list votes count:

   1. Any votes for non-winning local candidates are transferred entirely to their national list
   2. For the candidate winning a constituency, only the surplus votes are transferred to the national list.
      This is actually what your example represents; if we assume your totals to be the local constituency total then party A would receive 999 list votes (they needed 3001 to win the constituency)

   (The above is subject to the 5 % threshold you mentioned; I did not bother looking up whether the threshold applies to national list votes only or including the transferred votes. In most cases, it would not differ too significantly.)

Revisiting your example and assuming that each of the 106 constituencies voted exactly like that (meaning party A gains 106 seats of the constituencies) and also assuming that all over the country the national list votes are identical, my Excel file tells me that party A will win 142 seats (71 %) while party B wins 43 (22 %) and party C 14 (7 %). (Results like that are highly unlikely as rarely a party will win such a landslide all over the country; although Bavarian elections have resulted in all constituencies being held by CSU candidates.)

This is not a Mixed-Member Proportional representation (MMP) system but rather a Parallel Voting system. The outcome of the FPTP constituency seats has little to no bearing on how the national list seats are allocated. Similar systems exist e.g. in Japan, where the 465 seats in the House of Representatives are split into 289 single-member constituencies and 176 proportional representation seats. (The latter in Japan are not allocated nationwide but on a regional level.)

The Hungarian system includes a bonus to those parties who did less well in securing constituencies. In a pure FPTP system such as the British system, if one party scored 10000 votes in every constituency but the other 10001, the second would win all seats. In a simple Parallel Voting system as in Japan, they might also score 50 % of the proportional representation seats. In the Hungarian system, this effect is balanced out as the losing party can take 10000 votes per constituency into the national vote election but the winning party only 1, leading to a roughly 2:1 ratio in favour of the FPTP loser. This seems to strengthen the proportional aspect of the system.

However, the majority of seats are still allocated by FPTP constituencies so in essence securing relative majorities all across the country will likely remain more important.

The actual question was: Is this democratic?
The answer is a clear yes.

First of all, the formal requirements are present: every vote counts the same in the constituency it is cast in, everyone has (to the best of my knowledge) the right to vote and there is a real choice of candidates.

Second, the constituency winners are chosen by a clearly democratic process: first past the post. This is standard where only one position is open for election but where it is not considered important enough for an absolute majority (so a runoff election is not required). This also applies to the national list vote as the D’Hondt process, while systematically favouring large parties when it comes to residual seats, still represents a common quota system. (In most elections there are only minor differences between the results according to d’Hondt, Sainte-Laguë or other quota systems.)

The transfer of constituency losers votes (and surplus votes) actually makes the proportional block slightly more balancing imho, as not every vote on a losing candidate is lost. So those who voted for a candidate of the losing party are better represented in the national list vote which balances out the skewing of FPTP somewhat. But even without such a transfer (again: see Japan) this system would be considered democratic. And even without nationwide lists (see UK) this would be seen as democratic.

It is not a system of proportional representation like the one used in e.g. Germany. But there is no requirement for democratic elections to function according to proportional representation. All of these systems have their merits and disadvantages and deciding which one to use is an inherently political decision.