The two parties take very different approaches to the election of pledged delegates. In the Republican Party, the influence of winner-take-all states and winner-take-most states allows a front runner to win the nomination while only getting a plurality of the vote. In the Democratic Party, the fact that 14% of the delegates (officially unpledged party leader and elected officials, unofficially superdelegates) go to the convention as unpledged delegates and the pledged delegates are allocated proportionately, make it hard for even a clear front-runner with a majority of the votes to win enough pledged delegates unless the other candidates suspend their campaigns. As a result, for the second competitive cycle in a row, both candidates need the support of at least some of the super delegates to win the nomination.
There are a lot of different arguments for what superdelegates should consider in making their decision. The problem for Bernie Sanders and his supporters is that almost every argument favors Hillary Clinton.
In 2008, the winning argument for President Obama (and the argument that Sanders supporters were making in February) was that the supe delegates should go with whomever got the most pledged delegates. While there is still a slim chance that Sanders could catch Clinton in the pledged delegate count, it would take getting 60-65% in all of the remaining contests (plus some favorable breaks in the caucus states that wait until the state convention to allocate delegates).
In 2008, Clinton tried to make the argument that superdelegates should go with the total vote count, not the delegate count. The problem with the popular vote count is that certain caucus states (Iowa, Maine, Nevada, Washington, and Wyoming) don’t do a popular vote count. In any case, Clinton currently leads the popular vote count by 3 million votes; so the popular vote count also favors Clinton.
The next suggestion is that superdelegates should pay attention to their states. While this result is better for Sanders than the current pledges of super delegates, Clinton has won states that have a combined total of 379 super delegates, slightly more than half of the total number of superdelegates. So, at the end of the day, this argument does not work for Sanders either. (Modifying the argument so that members of Congress should go with whichever candidate won their district would benefit Sanders, but only net him thirteen more superdelegates than going with the state-wide results. That would still leave Clinton with 366 super delegates.) There are 156 superdelegates in the state won by Sanders (169 going by the district results) with 178 superdelegates in the remaining states. In short, this argument does not translate into enough superdelegates to overcome the gap in pledged delegates.
The next argument concerns the states that each candidate has won. However, this argument does not really help either candidate. Categorizing states is easier said than done because all line drawing is arbitrary. For the purposes of this analysis, I used the following definitions: Solid — a state has gone for the same party in all of the last four elections and by a margin over 12% in each of the last two elections; Likely — a state has gone for the same party in all of the last four elections and by a margin of over 8% in each of the last two elections; Lean — a state has gone for the same party in the last two elections and by a margin of over 6% in each of the last two elections; Swing — all other states.
By these terms, there are fifteen solid democratic states — Clinton has won six (all primaries), Sanders has won five (two primaries and three caucuses), and four primaries remain (with Clinton favored to win two or three). There is only one likely Democratic state — Michigan (won by Sanders). There are four lean Democratic states. Clinton won one (caucus), Sanders won two (one caucus, one primary), and one primary remains (in which Clinton is favored). These states combine for 233 electoral votes. In short, Clinton and Sanders are roughly even in the blue states.
There are eleven swing states, (some of these states are marginally swing but had at least one close election in the last two cycles even though they have gone with the same party in three of the last four elections.) Clinton won eight states (seven primaries, one caucus) to three for Sanders (two primaries, one caucus). The results in the swing states would indicate that super delegates should support Clinton.
There are no lean Republican states, and six likely Republican states. Clinton has won four of the six (all primaries) and tow remain (one primary, one caucus). A slight edge to Clinton.
While Sanders has tried to argue that Clinton is ahead because of her strong performance in the reddest states, the actual results show that this category is mixed. Clinton won the three biggest of the solid red states and one other solid red state (all in primaries). On the other hand, Sanders has won seven solid red states (one primary and six caucuses), but they were small states. Lastly, there are three solid red states still to vote. Again, the results from the solid red states are mixed.
That leaves the last argument — electability. The argument that Sanders is more electable is based on current polling for the November election. Polling at this stage of the race is questionable for two reasons: 1) candidates are not necessarily well-known at this stage of the race; 2) polling at this stage of the race can reflect temporary ill-feeling within a party that might not last much past the convention (effectively the anger stage of dealing with loss).
Traditionally, the Democrats have tended to pick candidates who are relatively unknown (Dukakis in 1988, Clinton in 1992, Obama in 2008). The unknown nature of the candidate has lended itself to volatility over the summer (against Dukakis in 1988 in favor of Clinton in 1992) as the voters got to know more about the candidates. With Trump and Clinton, the two candidates are well-enough known that the traditional volatility might be absent. With Sanders, there is room for movement as non-primary voters become familiar with him. Given the type of campaigns that Republicans have tended to run, there is good reason to believe that Sander’s current numbers represent a high water mark rather than what the numbers will look like in November.
Even taking the current numbers at full value, most of the recent polling shows Clinton solidly ahead of Trump nationally (the Real Clear Politics average has her up by 6.5%). However, that average includes a Rasmussen poll (who did not do a Sanders-Trump poll). Comparing only those pollsters who did polls for both Clinton and Sanders, Clinton leads Trump by 8.2% while Sanders leads Trump by 13.4%. While the better general election polling is a slight argument in Sanders’s favor, Clinton is still leading Trump by a substantial margin in the polling. Furthermore, when you look at recent state level polling, Clinton appears to have a safe lead in the solid/likely/lean Democratic states and is ahead in enough swing states to win the election by a comfortable margin. In short, if the role of the super delegates is to weed out unelectable candidates, Clinton seems to very electable. Even granting that Sanders might be more electable based on current polling, that slight edge would normally not be enough to convince super delegates to overturn the results of the primary. (If they were, Clinton might have won the nomination in 2008).
In short, barring some major event that changes the picture dramatically, there does not appear to be a solid argument for super delegates to support Sanders over Clinton. Given that Sanders needs to persuade a large number of superdelegates who are currently supporting Clinton to change their mind, the lack of a solid argument on his behalf will, in the end, likely prove fatal to any plan to win the nomination based on superdelegates switching to Sanders.