Objectives: The knowledge of the effects of white-collar crimes is incomplete. In the article, we operationalize white-collar crimes as bankruptcy frauds. Economic models maintain that interlinkages between firms may give ‘domino effects’: bankruptcy events could lead to ‘bankruptcy chains’ in which a bankruptcy spreads to other firms. Analogously, criminologists assert that social and economic networks can be a major source of fraud diffusion, with the potential to drive other firms bankrupt. Recent empirical results show that crimes may have detrimental and even asymmetric (nonlinear) effects on economic activity. We analyze the diffusion and the aggregate development of bankruptcy frauds in Sweden over nearly two hundred years, specifically focusing on the relationship between bankruptcy frauds and the bankruptcy volume. We also consider linkages between bankruptcy frauds, bankruptcies, and the macroeconomic cycle. Methods: We use long, aggregate time series, collected from several different historical and contemporary sources. Applying the recently developed cointegrating nonlinear autoregressive distributed lag (NARDL) model, we investigate whether the bankruptcy volume reacts asymmetrically to increases and decreases in bankruptcy frauds, both in the short and the long run. Results: Bankruptcy frauds reveal a causal effect on bankruptcies, showing an asymmetric (nonlinear) diffusion effect from economic frauds to the bankruptcy volume. Increases in bankruptcy frauds have a positive and significant effect on the bankruptcy volume. However, decreases in bankruptcy frauds show no significant effect. No causal relationship between the macroeconomic cycle and bankruptcy frauds is found. Conclusions: Our data and research approach demonstrate how previously generated hypotheses in both criminology and economic research on the relationship between (economic) crimes, economic activity, and the diffusion of white-collar crime can be tested at an aggregate level. © 2018 The Author(s)