As we learned in high school civics, the Founders failed to anticipate political parties when they designed a Constitution premised on a separation of powers into three separate branches. Policymaking power was delegated primarily to the legislative branch, not the executive branch. But Gary C. Jacobson’s new book, Presidents and Parties in the Public Mind, shows how legislators mostly rise and fall with a president of their own party, affecting their ability to legislate and win elections. With little or no change to the president’s formal powers, a government informally organized around parties has become a government informally organized around the president.

Relying mainly on survey data, Jacobson investigates the impact of the president on parties along a number of dimensions: public approval, coattails, polarization, and ideological placement. Party fortunes have depended on the president since the early republic, but Jacobson shows that the link has strengthened in recent decades. In many places, the book provides evidence for what observers already suspected about partisanship in a polarized era. But Jacobson’s careful statistical procedures and use of data offer the best evidence to date for the suspected patterns. He ultimately roots his findings in constructivist models of public opinion. Voters do not integrate their considerations into coherent worldviews that remain constant but base decisions on short-term memories where considerations are brought to mind through presidential actions and campaigns.

Even though partisans view presidents of their own party with rose-tinted glasses, presidents also shape voter evaluations of their parties. Respondents change not only their overall evaluations of the parties based on the president but also their assessment of what the parties stand for. Voters place parties on an ideological spectrum based on their perception of where a president or presidential candidate of that party lies on that spectrum. Voters even differentiate between competencies in various issue spheres, including the economy, minority rights, education, the environment, foreign policy, and terrorism. These assessments partially offset “issue ownership,” whereby voters as a whole tend to view one party as more competent than the other on an issue sphere on a semipermanent basis.

Party approval is not as volatile as approval of individual presidents, but it still moves with presidential approval. Today, presidential coattails affect down-ballot voting in presidential elections to an extent not seen since the nineteenth century, when voters could only choose one party ballot or another. As a result, America is functionally approaching a country with straight-ticket voting based on the presidential candidate.

Needless to say, the implications of this research are troubling. If legislators need their own party’s president to stay popular, they have a vested interest in the president’s success rather than checking and balancing potential overreach. Congressional leaders can also be at odds with presidents of their own party on a range of issues, as Speaker Tip O’Neill and President Jimmy Carter were. On economic issues, congressional parties (and earlier presidents and Congresses) are highly responsible for fiscal policy, but voters attribute economic well-being to the White House. Retrospective voting doles out, at best, “rough justice” to the parties.

Presidents have the most lasting influence on relatively young cohorts who have not amassed as many political judgments. Jacobson writes that Barack Obama’s most lasting legacy for the Democratic Party might have been generating a backlash, resulting in Donald Trump’s Republican Party and its dismal approval among younger cohorts. It will be interesting to see what positions Democratic primary voters in 2020 take in response to Trump. Jacobson’s work suggests that voters looking at their long-term policy prospects will have to consider not only a candidate’s policies but the impact that candidate will have on the perceptions of their favored party.