One thing is certain: If the Republican party had the same rules in 2015 as it did in 1915, Donald Trump would not be a problem. His entire candidacy is premised upon the breakdown of the relationship between the elites and the base, whose mutual accord has been central to the effectiveness of the postreform rules. For all its defects, the old nomination process, which placed power exclusively in the hands of the state and local organizations, would have dispatched Trump's candidacy with ruthless efficiency...
After this cycle, the Republican party desperately needs to reform its rules. This does not mean tinkering at the margins, playing with delegate allocation formulas and debate schedules, as the Republican National Committee did after 2012. The 1970s reforms allocated power within the party based on a premise of mutual trust and respect between the voters and the establishment. Without that foundation, the rules are a liability and need to be substantially redrafted. The real danger is not that a clownish demagogue like Trump will win the nomination this cycle, but that a demagogue who is not so much a clown eventually will.---Jay Cost, The Weekly Standard,
The mixture of democracy and free markets (and free movement of people) is unstable. This is due to the fact that markets are inherently disruptive, given that they aggregate the the constantly bombarded preferences of (thereby?) fickle consumers, and that (democratically established and/or enforced) existing rules skew the outcomes toward those with political clout. As I noted before this instability can become toxic if, in the context of heterogenous populations, the status quo generates reasonable expectations that get violated--as they must given the way markets generate uncertain outcomes. In times of stagnation, increasing inequality, and huge gains for well-protected (rent-seeking) insiders, it is to be expected that populist appeals can be mapped onto some us/them distinction that allows some us to be the authentic, united people -- even a moral community -- that excludes elites and would be (inauthentic -- e.g., immigrants, wrong religions, 'inferior' ethnicities, etc. --) members of the people.* The electoral and political success of populism, and (more important) accompanying polarization and destabilization of the familiar status quo, in turn makes some of the elites mistrustful of democracy.
One such mistrust, familiar from libertarian circles (recall) with mistrust of redistribution and rent-seeking, sees in democracy a potent moral danger in which in the polling booth the uneducated, whom cannot grasp complex phenomena, make bad decisions in the voting booth that turn out to have immoral consequences. Such libertarians prefer either markets or experts (or both, but with preference for markets) over democracy (sometimes they prefer lots as a species of democracy). Of course, it is not just libertarians that may prefer expert rule--as I have noted the European Union is a technocratic enterprise in which, thanks to the fate of Weimar and the instability of Third Republic, democratic 'input' is kept to a minimum.
There is also a longstanding conservative mistrust of democratic, popular sovereignty (going back to Burke). The impact of the Donald Trump is re-activating this trope the Republican party. What makes Jay Cost's analysis noteworthy is not that he is openly calling for an end to the accessible, primary system of picking the party-standard bearer, but rather his analysis that what makes intra-party democracy possible is a kind of tacit social compact involving mutual trust and respect between (party) 'base' and (party) 'elite.' This trust facilitates a structure in which "The [Republican] party voters possess the formal power to decide, but there is a vast infrastructure of donors, strategists, and insiders—an elite establishment—whose job is to control informally the people's decision." The intra-party success of populism (which is directed against the elite(s)) undermines that trust and respect (etc.).
As an aside, Goldwater-Reagan style conservativism was, itself, a successful, popular and (in its race-baiting moments, populist) revolt against moneyed, East coast elites (and the 'culture' of welfare/hand-outs, etc.). This conservative movement fizzled out, after forty to fifty years of success, in the second term of G.W. Bush (Iraq, Wall Street, etc.). (Trump is clearly a very different kind of populist.)
Cost, who calls for a return to a more heavy handed (formal) elite-managed selection procedure to maintain the integrity and longevity of the institution, does not reflect on the causes of the breakdown of trust between base and elite (which has been manifested, in part, by successful electoral challenges to (powerful) Republican incumbents). In fact, Cost is merely expressing the elite side of the breakdown of the trust. One important way in which trust breaks down if (tacitly) agreed upon expectations are systematically violated. (This is why there is a close connection between justice, trust, and stability.) The conservative movement has always been a coalition between cultural and Christian conservatives with (a much smaller group of) libertarians (who tended to represent, sometimes uneasily, business interests). While this is not always and everywhere a natural coalition, it is not entirely strange: markets and open borders generate permanent disruptions, and, for all the benefits these may bring, the subsequent dislocation feeds a desire for cultural and social stability (and slowing down of change).
Cost names a real danger: that a demagogue who is not so much a clown (but still a bit!) eventually will win the nomination and, thus, have a shot at the Presidency. He does not name the other danger (exhibited by his piece, but prevalent in Europe, too): that the "vast infrastructure of donors, strategists, and insiders" -- the rich, the highly educated, and the technocrats -- will give up on democracy as an end, and only see it as as a, dispensable, tool.
* I am reflecting here also insights of recent work on contemporary populism by Jan-Werner Mueller and my own colleague Sarah de Lange.