impact of legal counsel on outcomes for poor tenants in New York City's Housing Court: Results of a randomized experiment, The

Law & Society Review, 2001 by Seron, Carroll, Van Ryzin, Gregg, Frankel, Martin

The findings reported in Table 3 address the issue of delays or burdens on the Court, again looking only at the effect of initial assignment. Overall, the findings reported in Table 3 do not provide much evidence in support of the hypothesis that lawyers create inefficiencies for the court system. Though treatment cases sit on the docket for significantly more time (about 111 days) than control cases (approximately 82 days), treatment cases do not generate significantly more court appearances or motions than control cases. In fact, the findings actually suggest that lawyers may create some efficiencies for the Court. Treatment cases are significantly less likely to have post-judgment motions filed (approximately 13%) than control cases (29%). Post judgment motions may take one of three forms: (1) motions seeking a stay on eviction and extensions of time to pay arrears; (2) motions seeking to set aside the stipulation entered into, as it was unduly harsh or improvidently entered into; and (3) motions seeking to be restored to an apartment after eviction. Post-judgment motions are especially burdensome for the Court because they require a case to be reviewed and reopened after what was supposed to have been a final resolution of the dispute. These results suggest that counsel are effective in obtaining stipulations and compliance, and this in turn may produce significantly fewer of these motions when clients are represented.

In interpreting these findings, it is important to note again that the treatment category includes some cases (n = 59) in which the client did not in fact receive the services of a lawyer (see Table 1). It is reasonable to assume that the impact of a lawyer, that is the impact of the treatment on the treated, may be larger than that reported in Tables 2 and 3.

Thus, Table 4 reports the results of an instrumental variables (IV) estimation of the effect of legal representation itself on the same outcomes as in Table 2. The TV analysis adjusts for the fact that not all of those in the treatment group received the help of a lawyer, as well as the fact that a few of those in the control group found lawyers on their own. Thus these results can be interpreted as the effect of the treatment itself on the treated (Angrist et al. 1996).6 As the data in Table 4 show, the treatment effects on all five outcomes are highly significant statistically and of larger magnitude than the results shown in Table 2.

Table 5 similarly shows the IV results for the court process measures (Table 3). The difference in the mean number of court appearances is no longer statistically significant, as it was in Table 3. However, the difference in days is larger than in Table 3, and it remains significant statistically. Finally, the difference in post-- judgment motions remains significant and somewhat larger in magnitude than the results in Table 3.

Discussion and Conclusion

The findings from this experiment clearly show that when low-income tenants in New York City's Housing Court are provided with legal counsel, they experience significantly more beneficial procedural outcomes than their pro se counterparts. Represented tenants are much less likely to have a final judgment and order of eviction against them and more likely to benefit from a stipulation requiring a rent abatement or repair to their apartment. Because this evaluation is based on a true randomized experiment, these differences in outcomes can be attributed solely to the presence of legal counsel and are independent of the merits of the case. Moreover, these outcomes do not appear to come at much expense in terms of the efficiency of the Court; in fact, the presence of an attorney at the tenant's side may actually enhance efficiency by reducing the number of motions, particularly post-judgment motions.