Just to clarify things a little with Germany, and to add a more general perspective; the Huber model of confidence votes, as summarized by Döring and Hönnige (2005)
proceeds in four steps: Firstly, a cabinet minister proposes a policy. This policy can secondly either
be accepted by the majority or rejected. In the latter case the majority proposes a refined policy which is either
trying to avoid a vote of confidence or not. This refined policy can thirdly either be accepted or rejected by the
Prime Minister who can then decide whether to use a vote of confidence to push his policy through. If he uses
the vote of confidence, his majority has to decide in the fourth step to accept this policy or vote the Prime
Minister out of office to avoid this unwanted policy (Huber 1996: 273). [...]
The Huber model relies on the assumption that a vote of confidence is decided upon by the
simple majority of those voting (i.e. plurality). But we find three different voting rules in
European countries, each of them has different effects on the likelihood of the Prime Minister
to win a vote of confidence and the point he is able to choose ceteris paribus. The three voting
rules are:
(1) The simple majority of those voting must support the government (e.g. United Kingdom)
(2) The absolute majority of those voting must support the government (e.g. Spain)
(3) The government wins unless an absolute majority of the chambers members votes it down (e.g. France).
Huber states that Germany like France V is in the third category (1996: 271). This is not true.
According to Article 68 of the constitution the Chancellor must win the votes of the majority
of the members of the Bundestag, the so called Kanzlermehrheit, to survive a vote of
confidence. Thus, like Spain, Germany is in the second category needing an absolute
majority.
The difference of these voting rules lies in the way abstentions are counted and work in
favour or to the detriment of the Prime Minister. In the second category abstentions in the
governmental coalition work against the government, oppositional abstentions have no
effects. In the third category, abstentions on both sides work in favour of the government. In
the first category the effect of abstentions is dependent on number of governmental
abstentions versus abstentions on the opposition side. This leads to the conclusion that the
governments in category 3 have the highest incentive to use a vote of confidence and the
governments in category 2 have the lowest.
So Germany is roughly like the UK tradition in terms of article 68 of the Grundgesetz, but in a finer classification Spain and Germany belong in the same (separate) category and have somewhat different (tighter) rules for winning confidence votes.
With the caveat that this is pretty dated (and probably has some errors), here's Huber's (1996) summary table:
So there were (in 1996) at least a few other countries that had a "simple majority of those voting" confidence vote rule and that also a had a codified procedure either as laws, Standing Orders or a combination of both: Portugal, Italy, Denmark, Belgium and Ireland. The latter two didn't have any laws in this regard and relied only on Standing Orders.
But going back to Döring and Hönnige, another important issue is what happens if the government loses the confidence vote. Here we have another source of variation, which sets Germany aside:
The second major difference is the dissolution of parliament by the Prime Minister after he
has lost a vote of confidence. The basic feature of the confidence vote and the theoretical
models to understand it is the introduction of electoral cost in the considerations of members
of the majority parties. Members who wouldn’t agree with the Prime Ministers policy are
forced to accept it because he threatens them with dissolution of parliament, general elections
and thus a possible loss of their seat. In Diermeier/Feddersen's (1998) understanding they
choose between the control of a single policy and the control of all future policies.
Hence, the whole idea of the vote of confidence procedure is based on the dissolution threat.
This dissolution can be achieved by free will of the Prime Minister or enforced by tradition,
the standing orders of the parliament or the constitution. While this is true for most countries,
this is not true for Germany.
Indeed, two constitutional provisions define the situation here. The first one (Article 68) states
that the Chancellor may ask the Federal President to dissolve Parliament. After this question, the Federal President may dissolve parliament or he may not. So, the decision to dissolve the
Bundestag is not entirely in the hands of the Chancellor but in those of the President who in
case of a lost confidence vote of the Chancellor regains some powers he was stripped off after
the collapse of the Weimar Republic. In fact, a new veto player is introduced in the decision
to dissolve and we find a nested decision. According to the second provision (Article 67) a
chancellor can only be deselected by a constructive vote of no confidence which means, that a
new chancellor has to be elected by secret vote. A lost vote of confidence on its own does not
automatically lead to the resignation of the chancellor, dissolution of the parliament or the
formation of a new government. Indeed, a German chancellor might loose as many votes of
confidence as he likes to without any consequences (De Winter 1995).
There was a further development in Germany in 2005, not covered in that paper in a decision of the Constitutional Court. But this largely upheld the President's independent role:
the majority held that the decision of the German President regarding the
dissolution of the Parliament must serve as a legitimate check on the actions of the
Chancellor. The President ensures that the procedure has been taken in conformity
with the Constitution. If the procedure has been correct, the President must also
decide, in his or her discretion, whether the dissolution of the Parliament, the
shortening of the legislative period, and other political consequences are
proportional and whether he or she is willing to take the responsibility for them.
In the present case, the Court summarily concluded that the President had not
abused his discretion in approving the Chancellor’s requested dissolution.
It appears therefore that the Court, as in 1983, tried to square the circle: Both times a
Chancellor had apparently asked for a vote of confidence with the hope of losing,
merely in order to free the way for premature elections; both times the Federal
Constitutional Court held that the mere desire for new elections was not a sufficient
basis for the dissolution, yet in both cases it approved of the dissolutions in
question. Not surprisingly, this realpolitik decision found its critics – not only
outside, but also inside the Court.
