I think the bottom line is that the two systems, on the stability measures of length between elections and length of government, are actually quite similar. There's a tiny edge on these long-term longitudinal studies for PR systems, but it's not statistically significant. A lot of the studies that might have been taken into account back in the early 2000s were throwing Italy in. What were effectively cabinet shuffles—
On September 1st, 2016. See this statement in context.