The Pink Tide had a profound political and economic impact on Latin America. While the existing literature explores its drivers and reversal, this article addresses a gap by combining two decomposition methods with a dataset on values and perceptions. It makes three contributions. First, it quantifies the role of two key drivers: slow structural changes in society and faster shifts in political preferences. Second, it identifies which segments of the electorate experienced the largest preference shifts, using a novel combination of Oaxaca–Blinder and recentered influence function regressions—yielding the first empirical estimates of changes across the preference distribution. Third, it expands the use of these decomposition techniques beyond economic outcomes, offering tools for broader political analysis. The results show that 90 percent of the leftward shift during the Pink Tide stemmed from changes in voter preferences, not strategic “vote lending.” The entire ideological distribution moved leftward, not just centrist voters. Similarly, the rightward reversal was driven entirely by preference changes and extended across nearly the whole electorate. JEL Classification Codes: D72; O54; C21; C51; P16