My read of the literature is that strong presidential systems, such as we have in the United States, are moderately ill-advised. The tendency in such systems is for the President and the legislature to every now-and-again find themselves in an intractable disagreement—the scenario knows as “gridlock” in American politics—that doesn’t admit of any clear mechanism for resolution. This often leads to either the President mounting a coup (as used to happen a lot in Latin America) or else to Congress mounting a somewhat specious impeachment drive (as happened in the US during the Johnson administration and has tended to happen more recently in Latin America) in an effort to convert the regime to something more like a de facto parliamentary system.
Be all that as it may, strong presidential systems are also pretty rare outside of the Western Hemisphere where the US influence was very strong during the days of independence. In the twentieth century when the United States was helping countries (Germany, Japan, Austria) under military occupation or emerging from Communist rule (Poland, Czech Republic, Slovakia, Hungary) write constitutions even we didn’t advise them to adopt our system. Which led me to wonder how Afghanistan wound up with a strong president. My colleague Colin Cookman was kind enough to send me a not-online article by Barnett Rubin which explains:
The issue of governmental systems came into sharp relief at the CLJ as calls rang out for an up-or-down vote on presidentialism versus parliamentarism. Nearly all Pushtun delegates, joined by some members from other ethnic groups, came out for presidentialism. A bloc of non-Pushtun delegates, however, strongly supported a parliamentary system. Both sides made cases that mixed genuine public considerations with ethnopolitical ambitions. For Pushtuns and reformers, presidentialism provided a way for one of their own—everyone knew that the first incumbent would be Karzai—to emerge from the Bonn compromise with non-Pushtun armed factions as the popularly elected head of state. There would be no uncertainty about who held legitimate executive power in Kabul, and Washington would retain the benefit of having a clearly identifiable Afghan partner whom it would know well and indeed preferred. The largely non-Pushtun delegates who opposed presidentialism saw in it a risk of personal and ethnic dictatorship. A parliamentary system, they argued, would likely result in coalition governments that would be more representative and inclusive, safer from potential abuses of executive power, and hence more stable.
This seems like a really bad reason to favor a presidential system. The plurality ethnic group sees it as a way to entrench their power and, besides, everyone has a particular president in mind who they like. And Rubin’s account of the presidentialists’ more policy oriented case is also unpersuasive to me, “In his speech to the CLJ’s closing session, President Karzai cited post-1945 Italy and India since the Congress Party’s decline as negative examples.”
India seems like the relevant example, where a high level of diversity plus parliamentarism makes it difficult to assemble a stable coalition. Parliamentary government would have a similar problem in Afghanistan. But at the same time, shifting to presidentialism doesn’t actually make the underlying diversity and other social cleavages go away. Instead, it tends to ensure that presidential elections will become zero-sum contests of power between a Pashto candidate and a Tajik candidate in which both sides need to court the support of an Uzbek mass murderer and then deal with the inevitable ensuing legitimacy problems by informally assembling a coalition anyway.