The 'bozza Chiti' and the effective threshold

From repubblica.it comes information on the size of constituencies in the new electoral system proposed in the 'bozza Chiti'. They will be 'sub-provinciale', at most, 'provinciale'.

Let's assume conservatively that the constituencies will be equal to the number of provinces.

Let's also assume, as a simplifying measure, that the constituencies will all be given the same number of members of parliament to elect.

So, dividing the number of parliamentarians in the bozza Chiti - 400 - by 103, we get 3.88 MPs to be elected per constituency.

The effective threshold per constituency is therefore: 75% / (3.88 + 1) = 15.36%.

Now, the effective threshold at the constituency level is not a good estimate of the effective threshold at national level. Rein Taagepeera, in a 2002 Electoral Studies article, suggests that the nationwide threshold is equal to:

• effective constituency threshold / (square root of number of districts), or
• 15.36 / 1o3 ^ 0.5 = 15.36 / 10.15 = 1.51%

So, if anything, the proposed system is going to present a less formidable barrier to parties than the previous system, which had a legal threshold of 2% (plus largest coalition loser). Admittedly, these are back of the envelope calculations, and the distribution of seats between constituencies may (significantly) affect the results. But, if a (more) reductive electoral system is a good thing for Italy, then we'd better hope that the constituencies are more numerous than the one hundred and three provinces.