Table 2 Income assistance survival probabilities

From: Evaluating search and matching models using experimental data

British Columbia
Month	Simulated	Actual		Simulated	Actual
	Control	Control		Program	Program		Simulated	Actual
	Group	Group	p-value	Group	Group	p-value	Impact	Impact	p-value
6	0.904	0.895	0.384	0.809	0.836	0.030	−0.095	−0.058	0.026
12	0.817	0.787	0.033	0.643	0.616	0.110	−0.175	−0.171	0.881
18	0.739	0.714	0.107	0.581	0.543	0.026	−0.158	−0.171	0.572
24	0.668	0.669	0.922	0.525	0.507	0.287	−0.143	−0.162	0.400
30	0.604	0.618	0.390	0.475	0.476	0.956	−0.129	−0.142	0.561
36	0.546	0.569	0.171	0.429	0.436	0.675	−0.117	−0.133	0.501
42	0.493	0.511	0.316	0.388	0.410	0.198	−0.105	−0.101	0.853
48	0.446	0.482	0.033	0.351	0.410	0.062	−0.095	−0.073	0.817
53	0.410	0.428	0.299	0.322	0.343	0.195	−0.088	−0.084	0.883
1–53			0.364			0.333
New Brunswick
Month	Simulated	Actual		Simulated	Actual
	Control	Control		Program	Program		Simulated	Actual
	Group	Group	p-value	Group	Group	p-value	Impact	Impact	p-value
6	0.876	0.866	0.417	0.650	0.806	0.001	−0.226	−0.060	0.001
12	0.767	0.766	0.942	0.381	0.626	0.001	−0.386	−0.140	0.001
18	0.672	0.639	0.048	0.333	0.510	0.001	−0.338	−0.129	0.001
24	0.588	0.584	0.798	0.292	0.470	0.001	−0.296	−0.114	0.001
30	0.515	0.530	0.399	0.256	0.433	0.001	−0.260	−0.097	0.001
36	0.451	0.451	0.984	0.224	0.361	0.001	−0.227	−0.090	0.001
42	0.395	0.394	0.940	0.196	0.323	0.001	−0.199	−0.071	0.001
48	0.346	0.359	0.455	0.172	0.298	0.001	−0.174	−0.062	0.001
53	0.310	0.294	0.332	0.154	0.248	0.001	−0.156	−0.046	0.001
1–53			0.576			0.001

Note: p-values for single months are based on t-tests that the actual fraction still on IA minus the model prediction is equal to zero. For the all months test, we use a log-rank test that the actual (pooled) exits from IA are equal to the predicted (pooled) exits for the model and the data. The statistical tests treat the model predictions as constants

Back to article page