To Stay or Go? Consumer Bank Switching Behaviour After Government Interventions

Maaike Diepstraten, Carin van der Cruijsen

Research output: Working paperOther research output


We analyse whether and how individual savings and current accounts holders respond to government interventions at banks. We are the first to employ a difference-in-difference analysis, distinguish between a nationalisation and a capital injection, and separate between the two banking products. We find that the aggregate switching behaviour of consumers at intervened banks is similar
before and after the troubles and intervention. This holds for both type of interventions, both type of products, and for switching from and to the intervened bank. However, we show heterogeneity in consumer responses to government interventions, depending on the type of intervention and banking
product. For example, compared to consumers who trust the government, consumers with no or little trust are more likely to switch away from a bank after a nationalisation, relative to customers of the control bank. This holds for switching with the savings and current account. This highlights that trust in the government is an important prerequisite for a successful nationalisation. Second,
responses depend on consumers’ level of risk aversion. Risk averse current account holders at a nationalised bank are more likely to switch away than customers of the control bank. This result indicates that interventions can make consumers more aware of the troubles the intervened bank faces, and result in an outflow of consumers if a large share is risk averse.
Original languageEnglish
Place of PublicationAmsterdam
Publication statusPublished - 17 Mar 2017

Publication series

NameDNB Working Paper


  • consumer bank switching
  • bail-outs
  • capital injection
  • nationalisation
  • trust in the government
  • risk aversion


Dive into the research topics of 'To Stay or Go? Consumer Bank Switching Behaviour After Government Interventions'. Together they form a unique fingerprint.

Cite this