1. Lower turnout is helping Bernie Sanders. The entire decline in turnout in Democratic presidential nomination contests has been among moderates--liberal turnout is equal to the record breaking levels of 2008.
According to exit polls, an estimated 342,000 self-identified liberals participated in Iowa, New Hampshire and Nevada in 2016, vs 343,000 in 2008. By contrast, turnout among moderates and conservatives declined from an estimated 301,000 to 160,000, for a drop of nearly 47%.
The only way for turnout to go up at this point is for moderates to start voting more, and exit polls show that moderates skew toward Clinton. In Iowa, she won moderates by 23% when the state was tied as a whole. In Nevada, she won moderates by 22%, while Sanders won liberals by 5%. In New Hampshire, there was less difference, but Clinton still did 2% better among moderates than she did among liberals.
2. But don't worry Sanders fans, there is zero chance turnout will increase in future primaries and caucuses. Following his defeat in the Nevada caucuses, Bernie Sanders blamed his defeat on low turnout, saying the following on Meet the Press:
...we will do well when young people, when working-class people come out. We do not do well when the voter turnout is not large. We did not do as good a job as I had wanted to bring out a large turnout.
Unfortunately for Senator Sanders, there is absolutely no chance of voter turnout going up in future contests. The high-attention early states in presidential nominating contests always have much high turnout rates than contests which take place later in the primary calendar. For example, in 2008, New Hampshire had a 54% turnout rate, with all other primaries averaging 30%. Iowa had a 16% turnout rate, with all other caucuses averaging 5%.
The reason for this should be obvious--campaigns lavish early states with months worth of personal appearances, dozens of field offices, and millions of dollars in paid advertising. Resources spent on later states are comparatively small.
However, as was noted in point #1, this isn't actually bad news for Sanders. Lower turnout is helping him so far, because 100% of the lower turnout is coming from self-identified moderates--a demographic that is pro-Clinton.
3. Primary turnout is not predictive of general election turnout. A concern expressed by many Democrats have right now is that while Democratic turnout is down from 2008, Republican primary turnout is at record breaking levels. While this is true, it is also true that the Republican turnout advantage over Democrats in early presidential nomination contests was much larger in 2000 than it so far has been in 2016, and Al Gore went on to win the popular vote in the general election anyway. The point being that there is no demonstrated connection between presidential primary turnout rates and presidential general election turnout rates.
4. The difference between high turnout and low turnout presidential elections is a lot less than you might think. In 2000--the second lowest turnout general election in history--turnout among the voting eligible population was 54.2%. In 2008--the highest turnout general election since 18-year olds were granted the franchise--turnout among the voting eligible population was 61.6%.
That 7.2% difference is the entire amount that voter turnout could potentially vary in 2016 (I'm basing this on two pretty safe assumptions--that the open Supreme Court seat makes equalling the lowest possible turnout of all-time impossible, and that surpassing the modern record set in 2008 is also impossible).
No matter who the candidates are, no matter how populist the message is, no matter how many volunteers, or how good the data or how clever the tactics, the absolute maximum range in voter turnout is 7.2% of the voting eligible population. In all likelihood, the range that is on the table in 2016 is less than 7.2%. Further, a not insignificant percentage of the unlikely voters who do turnout will go Republican, especially if the nominee is Trump.
"Turning out new voters" is not a magic wand, and it is not infinity. Lots of people have care enough about politics that they turn out in every presidential election, and a lot of other people will never turnout no matter what you do. Only a fairly small slice of the population legitimately wavers between voting and not voting in presidential elections--and not all of those people are Democrats or progressives. (Most of them probably don't even follow politics much at all, to be honest.)
Voter turnout efforts are absolutely worth pursuing, but they are difficult.
In 2014, Daily Kos teamed up with MoveOn to help turn out voters in crucial Senate races. The over 6,000 Daily Kos volunteers who participated in that program made 1.49 million phone calls to "drop-off" voters--that is, people who turn out for presidential elections but tend to stay home during midterms.Those 1.49 million phone calls made by the 6,000 volunteers generated 21,000 pledges to vote. That comes to 3.5 votes per volunteer, and one voter for every 700 phone calls made.
At that rate, turning out 1,400,000 unlikely voters--or, 1% of the 2016 electorate--would require 400,000 people making one billion phone calls.
I am the Executive Campaign Director here at Daily Kos, so you better believe that I am going try and make as big a dent through our GOTV efforts as we possibly can. But I mean, one billion phone calls? To increase voter turnout by just 1%? This is a difficult job, and we need to have our eyes open about just how difficult it is.