The COVID-19 pandemic and its associated lockdowns presented an unprecedented disruption to daily life, impacting mental health and behaviors significantly. Music listening, a common emotion regulation strategy, became a focus of research during this time. While surveys indicated increased nostalgic music listening, this study leverages the large dataset of Spotify streaming data to quantitatively analyze these behaviors in the UK during the first lockdown. The study aims to empirically determine if the lockdown caused a shift in music preference towards older songs and to investigate the relationship between this shift and the reported positivity bias in music choices during the pandemic.

Existing literature shows that negative emotions are often associated with nostalgia, which itself can have positive effects. Studies using surveys during the pandemic reported increased nostalgic music listening, sometimes linked to distress, and a general trend towards listening to happier music. However, reliance on self-reported data and lack of pre-pandemic benchmarks limit the conclusions of previous studies. This study aims to fill this gap by examining objective music streaming data.

The study uses Spotify's publicly available data on UK daily top 200 songs from January to July 2020, encompassing the first lockdown period. Songs older than five years were classified as 'old' music, while others were classified as 'recent'. A logistic regression analysis was employed to determine whether the lockdown (and the COVID-19 infection rate) significantly influenced the choice of 'old' versus 'recent' music. The regression includes control variables like the number of new releases, total daily plays, and the average 'oldness' in 2019. Robustness checks were performed using alternative lockdown start dates and comparing the data to the same period in 2019. Further analysis, using time series regression, examined the changes in the number of plays of positive and negative old and recent songs to investigate the relationship between nostalgia and positivity.

The logistic regression analysis showed a significant positive effect of the lockdown on listening to older music. Robustness checks confirmed that this change was not an annual pattern. Further analysis focusing on positive and negative songs indicated that while there was an increase in listening to older songs regardless of emotional valence, the surge in positive old music was more persistent than that of recent positive music. Analysis of audio features revealed that older songs in the sample had significantly higher average valence (positivity) scores than recent songs. Time series regression showed a persistent increase in listening to positive old music after the lockdown, but a short-lived increase followed by a decrease in recent positive music. Listening to old negative songs also experienced an increase, but with no sustained upward trend after the initial spike.

The findings suggest that the first COVID-19 lockdown in the UK led to a significant shift in music preferences toward older songs, likely driven by a need for nostalgia as an emotion regulation strategy. While a positivity bias in music choices was also observed, the study demonstrates that nostalgia-seeking behavior operated independently and even reinforced the preference for positive music. The sustained increase in positive old music suggests that nostalgia may offer a more enduring sense of comfort and emotional regulation than recent music, even if both cater to the need for positivity during stressful times.

This study provides robust evidence from Spotify streaming data supporting the hypothesis that the COVID-19 lockdown increased nostalgic music listening in the UK. The results suggest that nostalgia-seeking is a significant factor in music choice during times of stress, independent of but potentially reinforced by a preference for positive music. Future research could explore the psychological mechanisms linking nostalgia, music choice, and emotional well-being further and investigate how these patterns differ across various cultures and populations.

The study relies solely on Spotify data, which limits its ability to explore individual listener preferences or the emotional states associated with music choices. The definition of 'old' music (songs older than five years) is arbitrary and might not perfectly capture all forms of nostalgic listening. Finally, the analysis focuses on the UK and may not be generalizable to other cultural contexts.