As Ezra noted Thursday morning, the senators who blocked the Manchin-Toomey background checks bill represent about a third of the country's population.
And that's not even the worst it's gotten. In 1900, the largest state, New York, was a shocking 171.7 times the size of the smallest, Nevada. An individual Nevadan had 171.7 times the influence of New Yorkers in the upper house of Congress:
Another way to look at this is to ask what is the smallest population a majority of senators could represent in a given year. The way to calculate that is to divide the states in half by population, sum the populations of the smallest half, add the next largest state (so you have a majority, not just half; with an odd number of states you can just round up from the half figure), and then see what proportion of the total U.S. population that is.
For example, in 1790 we had 13 states. If all the senators from the seven smallest states voted the same way, they'd have a majority. Those seven states' populations, however, accounted for 27.4 percent of the population. With every state we add, the proportion of the population needed to pass stuff through the Senate has, generally, fallen, until it reached its current point of 17.82 percent:
That's right: If senators representing 17.82 percent of the population agree, they can get a majority in the 2013 U.S. Senate. That's not the lowest that figure has gotten (it hit about 16.8 percent in 1970) but it's about there. And this doesn't even take the filibuster into account. The smallest 20 states amount to 11.27 percent of the U.S. population, but if all of their senators band together they can successfully filibuster legislation.
You can see all the data represented here in this spreadsheet.
Update - The second graph's 1790 calculations were off, and have been corrected. We regret the error.