Investigating causality among unemployment, income and crime in Taiwan: evidence from the bounds test approach

Abstract

This paper examines the long-run and causal relationships among unemployment, income and crime in Taiwan. The results clearly indicate that there is a long-run level equilibrium relationship among unemployment, income and total crime. There are also long-run relationships among unemployment, income and theft and among unemployment, income and economic fraud. The causality test results from the ECM-VAR and level VAR models indicate that there is a neutral relationship among unemployment, income and total crime, and a neutral relationship among unemployment, income and all three categories of crime. It is concluded that there is no strong evidence in favor of the unemployment-led crime (ULC) or the crime-led unemployment (CLU) hypotheses in Taiwan.

Keywords:

• crime
• unemployment, bounds test
• Granger causality

JEL Classifications:

• C32
• J60

Acknowledgements

We would like to thank the editor Professor Xiaming Liu and two anonymous referees of this journal for helpful comments and suggestions. Financial support from the National Science Council (NSC 95-2415-H-212-001) is gratefully acknowledged. The usual disclaimer applies.

Notes

Notes

1. Readers are referred to Pyle (1995) or Ehrlich (1996) for a review of the literature before 1990. Narayan and Smyth (2004) also provide an excellent literature review on the relationship between crime and unemployment.

2. In 1990, the stock index plunged from 12,000 to 2000 points, and many investors lost almost everything.

3. Narayan and Smyth (2004) discuss the advantage of the bounds test approach, reporting that ‘policy-makers can determine which type of crime is statistically endogenous and which is exogenous and distinguish between “short-term” and “long-term” Granger causality.’

4. Turner (2006) recently generated critical values based on the response surfaces of an F-test for cointegration.

5. We abbreviate average monthly earnings in manufacturing as ‘income’ in the text that follows.

6. The national policy agency defines ‘total theft’ as ‘petty theft, common theft, car theft and motorcycle theft’. Total crime is the sum of ‘total theft’ plus ‘total violence’. ‘Total violence’ is ‘first degree murder, kidnapping and bribery, robbery, aggravated assault, threatening, aggravated burglary, sexual assault and rape’.

7. Masih and Masih (1996) consider some socioeconomic determinants, e.g. rates of urbanization, police strength and housing starts, in testing the unemployment–crime causal nexus for Australia. We do not consider these determinants in our empirical study because of lack of data for Taiwan.

8. Thanks are due to an anonymous referee for pointing this out to us.

9. Engle and Granger (1987) demonstrated that once a number of variables are found to be cointegrated, there always exists a corresponding error correction representation which implies that changes in the dependent variable are a function of the level of disequilibrium in the cointegrating relationship (captured by the error correction term) as well as changes in other explanatory variable(s). Through the error correction term, the ECM opens up an additional channel for Granger causality.