The Referendum That Held

Italy just watched that move fail.

On June 25 and 26, voters rejected a sweeping package of constitutional changes by 61.3 percent to 38.7 percent. The package was the signature legacy project of the Berlusconi government — drafted, passed through parliament on a bare partisan majority in 2005, and left waiting for the confirmatory referendum the constitution requires when reform passes without a two-thirds supermajority. By the time the ballots were counted, the government that wrote it was already gone, turned out in April by Romano Prodi's center-left. So this was a departing administration asking the electorate to make its preferences permanent after the electorate had already shown it the door. The voters declined to be governed posthumously.

Read the brochure and the reform was "modernization": fewer deputies, devolved powers to the regions, a streamlined federal Senate, a Prime Minister recast as a directly empowered "Premier." Strip the brochure language and you get the actual mechanism — a head of government who could call for parliament's dissolution, and whom parliament could no longer remove without triggering its own destruction. That is not modernization. That is a one-way valve. The reform took the most basic brake in a parliamentary system, the chamber's power to unseat the executive, and quietly cut the cable.

This is the part worth naming, because the coverage treats it as an Italian story and it is not. It is a recurring subroutine. A government reaches the end of its leverage and reframes the concentration of its own power as constitutional housekeeping. Napoleon III ran the same play with plebiscites, dressing a coup in the language of popular will. De Gaulle reached for direct ratification one time too many in 1969, asked the country to bless a restructuring of the Senate and regions, lost, and resigned the next morning. The names rotate. The rhetoric gets a fresh coat. The shape underneath is identical: ask the population to ratify the removal of the thing that can stop you.

What makes the Italian result interesting is not that the move was attempted — it always is — but where it broke. The institutions had already bent. Parliament passed the package. The reform cleared every gate the political class controls. The only mechanism left standing was the one the political class does not own: the population, voting directly, on a single question, with no coalition to whip and no committee to bury it. The brake that held was the one nobody could capture in advance.

There is a comforting story you could tell here, about the wisdom of the people and the resilience of democracy. Resist it. Direct ratification is not a reliable brake; it is a coin the powerful keep flipping because it pays off more often than it fails. Napoleon III built an empire on it — his plebiscites delivered crushing yes majorities on demand, a coup and then a crown ratified by overwhelming popular vote. The machinery of a managed yes is old and well-tooled. What we remember are the misfires: De Gaulle in 1969, Pinochet in 1988, Italy now. That asymmetry is itself the tell. The move is attempted constantly and recalled mainly when it breaks, which means the failures feel like the rule when they are actually the exception. Italy's electorate read the fine print this time. There is no guarantee the next electorate will, anywhere, when the package is branded better and the question is buried under three others designed to confuse the count.

So file 2006 as a data point, not a verdict. A government on its way out asked a country to make its power structural, and the country said no. The valuable lesson is not that the people are wise. It is that the move has a tell — every one of these reforms sells a concentration of power as an act of efficiency — and that the brake of last resort is whatever part of the system the people writing the reform forgot to own. And that brake works only while it can still see: while the population can still read past the brochure to the mechanism underneath. The vote is not the safeguard. The legibility is.

They will run it again. They always do. The only variable is whether the population still recognizes the brochure.