The paper, from which these studies were taken, considers new methods to examine voter participation (Study #2) and efforts to get-out-the-vote (Study #5). The studies confirm some previous work; however, these methods may provide new insight for those examining those particular areas.

Study #2
Regular Voters, Marginal Voters, and the Electoral Effects of Higher Turnout

Abstract How do marginal voters differ from regular voters? I develop a method for comparing the partisan preferences of regular voters to those marginal voters whose turnout decisions are influenced by exogenous factors and apply it to three sources of variation in turnout – weather and the timing of gubernatorial and congressional elections. In each setting, marginal voters are more supportive of the Democratic Party than regular voters, and the substantive size of this divide can be huge – ranging from 5 to 47 percentage points. The findings suggest that electoral reforms and other factors that may expand or contract the electorate can have important electoral consequences. Moreover, the findings suggest that election results do not always reflect the preferences of the citizenry, because those marginal citizens who stay home have systematically different preferences than those who participate.¹

From the conclusion (p. 61):

Every test in this paper points in the same direction. Regardless of the setting or the particular sample of marginal voters, regular voters are not representative of the larger pool of possible voters. Citizens on the margins are systematically more supportive of the Democratic Party than regular voters, and this gap can have significant electoral consequences. For example, the party of many states’ governors would be different if their elections were held in different years when a different subset of citizens turns out. Even if the introduction of marginal voters in the electorate does not change discrete electoral outcomes, the new composition of voters may still influence the platforms of candidates and the distribution of public services. Currently, American elections fail to reflect the preferences of all citizens because those on the margins are systematically different from those regularly participating. The repeated testing of preference gaps between marginal and regular voters may improve our understanding of this phenomenon and identify solutions for the mitigation of this participatory inequality. [Emboldening added.]

This conclusion suggests that the "interests of the nation" (the people as a whole) is not represented by those who vote.

Study #5
Increasing Inequality: The Effect of GOTV Mobilization on the Composition of the Electorate

Abstract
Numerous get-out-the-vote (GOTV) interventions are successful in raising voter turnout. However, these increases may not be evenly distributed across the electorate and may actually increase the differences between voters and non-voters. This phenomenon is particularly notable given the many GOTV strategies that explicitly aim to reduce inequalities in representation. By analyzing individual level-data, we reassess previous GOTV experiments to determine which interventions mobilize under-represented versus well-represented citizens. We develop a generalized and exportable test which indicates whether a particular intervention reduces or exacerbates disparities in political participation and apply it to 26 previous experimental interventions. Despite raising mean levels of voter turnout, more than two-thirds of the interventions in our sample widened disparities in participation. On average, voter mobilization strategies tend to increase the participation gap, thereby exacerbating representational inequality. We conclude by discussing substantive implications for political representation and methodological implications for experimenters.¹

The above figure (p. 146) presents the results of selected experimental interventions, thus is not conclusive for all GOTV campaigns. "Propensity" (x-axis) is scaled for standard deviation, where the values are from -2 (less likely to vote) through +2 (more likely to vote). The y-axis indicates the effect (increase in voter participation) for the various experiments.

The blue lines show a reduction in the voter participation gap.

Only two interventions in our analysis demonstrate statistically significant evidence that the participation gap was reduced. What might explain the difference in these two cases? One intriguing similarity between the two experiments with negative interaction effects is that they both targeted citizens in communities with large African American populations. One explicitly targeted African Americans (Middleton and Green 2008) and the other was set in the largely African American city of Detroit (Gerber, Green, and Nickerson 2003). (p.179)

The red lines (more typical of GOTV campaigns) show an increase in the voter participation gap; thus those who are not as well represented in government, become even less so. This does not mean there will be no change in government from the GOTV campaign, only that there is no assurance that such a change will increase representation for those who don't participate as voters.

From the conclusion (p. 152):

The findings of this paper also raise an ethical concern for experimenters and practitioners because experimental interventions and mobilization efforts are often conducted with the assumption that raising average participation levels can only be good for democracy. However, the evidence in this paper – that voter mobilization tends to exacerbate existing inequalities in the electorate – necessitates a more nuanced perspective. Despite good intentions, current GOTV efforts are not the solution to persistent inequalities in the political process. On the contrary, these efforts may contribute to the problem by making the electorate more polarized and less representative of the greater population.

¹ Fowler, Anthony George. 2013. Five Studies on the Causes and Consequences of Voter Turnout. Doctoral dissertation, Harvard University.