Divorce increased throughout the twentieth century across Western nations. Procyclical divorce, i.e. following macroeconomic cycles, is often suggested and women’s increased economic independence is considered a contributor to increasing divorce rates. Research has, however, overlooked that different factors may drive divorce rates across time horizons. Macroeconomic conditions, like unemployment, covary with divorce in the short term, while processes like women’s economic independence, correlate with the long-term divorce trend. Most studies focus on post-war United States, neglecting historical perspectives and Nordic countries that pioneered high divorce rates. To address these gaps, we used band spectrum regression to analyse time series capturing macroeconomic conditions, women’s economic independence, welfare state expansion, and divorce rates in Sweden, 1915–2010. This method differentiates between short- and long-run associations. Results show that this is important because most variables only covary with the divorce rate within one horizon. Results show that context matter. Contrasting U.S. findings, including novel estimates for 1930–2010, our results suggest countercyclical divorce and positive correlations between government expenditure, female-to-male relative wages, and the divorce trend in Sweden. This likely reflects the stabilising effect of the growing welfare state, which lessened financial barriers to divorce and promoted economic independence through female employment, transfers and services.