Eugene Volokh today talks about the 17th Amendment (while Jacob T. Levy links to a September post of his discussing the same topic). The 17th Amendment, adopted 80 years ago, changed the system for election of senators to the current system of popular election; originally, they were elected by the state legislatures (Article I, Section 3, Paragraph 1). The indirect election of Senators was primarily designed to give the states direct input in federal policymaking (see, for example, Federalist 62), and thus in some sense to reinforce federalism by acting as a check on the House of Representatives deciding to exceed the enumerated powers of Congress. (Some argue that the 17th Amendment, rather than the 16th, was the true enabler for big government.)

Of course, there are other issues at work that prevented the Senate from being an effective bulwark on federal power (even before the 17th Amendment); the principal-agent problem loomed large, with the interests of senators not necessarily coinciding with the state legislatures that chose them. One reason why the German Bundesrat works and persists as a powerful, indirectly elected chamber is that its members are true agents of the Land's parliament; by contrast, the pre-17th Senate had very weak principal-agent links (in order to promote institutional stability, another goal of the Senate's design), although whether the Framers intended it that way is unclear.

Jacob writes:

Even if the original 1787 apparatus were clearly better as a matter of constitutional engineering than the current mechanism, it might have been too politically fragile. If it had not bent with the 17th amendment, it might have broken later say, during the Terrible Twenties and Thirties when constitutional democracies were swept away by populist-authoritarianism in much of the world, and we had Longs, Coughlins, and Roosevelts of our own. A defense of the 17th along these lines is kind of like a defense of the 1937 "switch in time that saved nine" by the Supreme Court, not because the switch was constitutionally correct, but because it did manage to "save nine"-- that is, to save substantial judicial independence from a court packing precedent that would have left us with New Deal constitutional revisionism and with a cowed, subservient judiciary and with a precedent for presidents changing the constitutional rules whenever they weren't getting their way.

On balance, I think that the 17th Amendment had little effect in the long term. By the time of its passage, over half the states already had a de facto system for electing senators by popular vote; doubtless most other states would have followed suit eventually. In other words, the 17th is the result of a trend that was taking place anyway, codifying (and legitimating) the Progressivist change that was already on its way to being implemented nationwide. (And, among those changes, it was probably the least harmful procedural change that could have been institutionalized in the Constitution; far more harmful would have been a constitutionally-blessed role for parties as public utilities or a mandate for open primary elections. Of course, in other realms, the 18th was orders of magnitude worse, and also the result of Progressivist do-gooderism.) Ultimately the 17th was no more than a small chink in the armor of federalism, but if the Senate had remained a truly appointed body into the 1930s I agree its legitimacy would have been called into serious question — and perhaps its powers would have been evicerated by a Supreme Court more interested in preserving Marbury v. Madison and judicial review than federalism.

Incidentally, the relative design of the European Council of Ministers and the European Parliament is similar to that of the Bundesrat and Bundestag, although the EU institutions may be the sole current example of asymmetrical bicameralism where the upper chamber is vastly more powerful than the lower; however, it is telling that the phrase “democratic deficit” is virtually synonymous with the EU.