Criminal defendants' satisfaction with lawyers: perceptions of procedural fairness and effort of the lawyer

Abstract

At present, procedural justice theory has predominantly been used to explain defendants' satisfaction with the police, courts and prisons. It is unclear to what extent this theory is also applicable to lawyers. This study investigates to what extent (1) criminal defendants are satisfied with their lawyers and (2) procedural fairness characteristics and the effort of the lawyer are related to defendants' satisfaction with their lawyers. Data from the Prison Project were used: a large-scale research project among Dutch criminal defendants (N = 1479). Results suggest that generally, Dutch defendants are very satisfied with their lawyers. Variation in defendants' satisfaction with their lawyers can be attributed for a substantial part to procedural fairness characteristics.

Keywords:

• criminal defendants
• lawyers
• perceptions
• procedural justice
• satisfaction

Acknowledgements

The Prison Project is financially supported by the University of Leiden, the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), the Netherlands Organization for Scientific Research (NWO) and Utrecht University.

Notes

1. We created a dummy variable for missing/unknown data for ethnicity as this variable contained a relatively large number of missing values.

2. This method uses the mean of the available items to measure the same construct (Roth, Switzer, & Switzer, 1999). Hawthorne and Elliot (2005) demonstrated that this method outperformes listwise deletion even when more than 50% of the items were missing (cf. Miller-Clayton, 2010). In the current study, we used this method for the procedural fairness scale and the attitudes towards the law scale. We included cases if they had a non-missing value on at least 60% of the items.

3. This was mainly due to missing values on the procedural fairness scale (N = 1655), the number of times the defendant had been represented by a lawyer (N = 1685) and attitudes towards the law (N = 1664).

4. As the dependent variable is ordinal, categorical data analyses were performed as well. Multinomial logistic regression analyses did not result in different findings.

5. Initially, we also included the item ‘My lawyer makes decisions based upon facts rather than opinions’, as was also done by Sunshine and Tyler (2003). However, including this item in the scale significantly decreased the Cronbach's alpha to .55.

6. If defendants were accused of multiple offences, they were coded according to the most serious offence they were accused of.

7. As the dependent variable is distributed rather unevenly, logistic regression analyses were employed as well (values 1–3 coded as ‘0’, values 4–5 coded as ‘1’), but this did not result in different findings.

8. We also dichotomized the variables at 25%, but this only resulted in different p values for frequency of visit at the police office. On average, defendants were more satisfied with lawyers who had not visited them at the police office (M = 4.03, SD = 1.07) than with lawyers who had visited them one to six times at the police office (M = 4.01, SD = 1.00). This difference was not significant (t = .33, p >.05).

9. We also compared the satisfaction for defendants who score low or high on the independent variables by dichotomizing the dependent variable satisfaction. The values 1–3 were coded as ‘0’, while the values 4–5 were coded as ‘1’. This did not result in different p values.

10. We also performed our analyses using listwise deletion for all variables (N = 1370). Using dummy variables for missing/unknown data and the personal mean substitution method did not result in substantive differences in our findings.

11. In the interest of space and presentability, the dummy variables capturing the missing/unknown data are not presented in Table 3. Additional information on these estimates are available from the authors.

12. The variance inflation factors (VIF) and the tolerance statistics indicated no problems with multicollinearity as none of the tolerances were lower than .2, and none of the VIF were higher than 4.

13. Given the fact that defendants, who were represented by a private lawyer and who were represented more often before by their current lawyer, were more satisfied with their lawyer, we assessed potential interactions between these variables and other significant independent variables (i.e. procedural fairness and the number of phone calls). Separate analyses revealed that the more often defendants were represented before by their current lawyer, the weaker the relation between procedural fairness and satisfaction with the lawyer. In addition, for defendants who were represented by a private lawyer, the relation between procedural fairness and satisfaction with the lawyer was weaker. However, the standardized coefficients for both interaction terms were too small to be considered substantive (β ≤ .05). Moreover, none of the remaining interaction effects proved significant.