[Carlo Lazzaro]

Sunil:
I overlooked that you imposed cluster-robust standard errors in your regression model, whereas -hausman- allows default standard errors only.
Hence, you should test -fe- vs -re- specification with the community -contributed command -xtoverid- which, in turns, does not support -fvvarlist- notation, forcing you to prefix your regression model with -xi:- and/or creating interaction by hand.
I would run -xtoverid- to check whether -fe- is actually the way to go, first.

[Clyde Schechter]

The reason your interaction term is omitted due to colinearity is that your data are incorrectly structured. In the Treat = 0 group, the variable Entry fails to distinguish between the pre- and post- intervention years. In your data, Entry is set to 1 for all years in that group. This is wrong and is the source of the problem you report.

But, looking a bit more closely at your data, it appears that different fmid's make the transition from Entry = 0 to Entry = 1 in different years. So our data are not suitable for a classical difference in differences analysis. You must go instead to a generalized difference in difference analysis. In this analysis, the variables Entry and Treat play no role. Only the interaction term, EntryTreat will be used. The code will be something like this:

xtreg outcome i.EntryTreat i.year, fe

The inclusion of i.year is absolutely mandatory here. The generalized DID estimate of your treatment effect will be the coefficient of EntryTreat in the -xtreg- output.

With that out of the way, let me also reinforce Carlo's observation that the data set you posted was not as helpful as it needed to be: you are asking people to help you troubleshoot a regression, but you post a data set in which nearly all of the variables in the regression are not shown (and the data set is cluttered up with other variables that are not part of the regression and have no apparent relevance to your problem). To make the most out of a forum like this, it is important to show the data that goes along with the code you are asking for help with.

[Sunil Dhoubhadel]

Clyde,

My apologies for not being able to properly format the data. I understand the anxiety created by this. Per your suggestion, I used dataex output and when I cut and paste the data it messed up the format. I have posted below the formatted data for all the variables that I used in my model for you to look at.

In my model, Entry=1 codes the pre-treatment years and Treat =1 codes for the treated farms. As you have correctly noted for Treat =0 farms, it fails to distinguish between pre-treatment and post-treatment years because they are never treated i.e. there is no post-treatment years for these farms. Also, you have correctly noted that all the treated farms are not treated in the same year, i.e. there is a staggered entry of treatment for the treated farms.

I want to test if there is any difference in outcome in pre-treatment years for both treated and untreated farms. If I just use the interaction term EntryTreat as you have suggested, it will fail to test the difference in outcome as I intended because EntryTreat =0, would include treated farms in post-treatment years as well as untreated farms in pre-treatment years as shown in the data below. So, if I correctly understood your suggestion, I am not sure just using the interaction term would serve my purpose. Please let me know if I am missing anything here. More clarification is highly appreciated.

Regards,
Sunil

fmid	year	fmsz	dbassra	insexp	age	prcrp	prown	mcrpyd	diff_nincpac	Entry	Treat	EntryTreat
11402300	1982	2747			62	62.24973	0			1	1	1
11402300	1983	1784			30	61.79372	0		18.38133	1	1	1
11402300	1984	1697			28	100	15.08544		24.7158	1	1	1
11402300	1985	1379			30	100.029	0		-53.2294	1	1	1
11402300	1986	1243			30	100.0483	0		20.54122	1	1	1
11402300	1987	1215			35	100	0		24.13819	1	1	1
11402300	1988	3401			30	52.66098	0		-20.803	1	1	1
11402300	1989	5156			33	33.87898	0		-23.6661	1	1	1
11402300	1990	3388				52.47934	0		36.44645	1	1	1
11402300	1991	2945			38	52.46519	0		-30.1498	1	1	1
11402300	1995	2888			42	67.86012	0			1	1	1
11402300	1996	3068			43	69.7425	0		90.96358	1	1	1
11402300	1997	2963			44	68.66014	0		-42.6524	1	1	1
11402300	1998	2795			45	66.85509	0		-38.9077	1	1	1
11402300	1999	2809			46	67.00249	0		14.35262	1	1	1
11402300	2000	2808			47	66.99786	0		-11.3067	1	1	1
11402300	2001	2634			48	64.84055	0		7.479562	1	1	1
11402300	2002	4262	0.84	8169	49	48.43735	0		-23.6915	1	1	1
11402300	2003	2821	0.98	3775	50	64.11201	0		54.61241	1	1	1
11402300	2004	2408	0.83	0	51	81.6902	0		-26.3217	1	1	1
11402300	2005	3188	0.85	17762	52	61.74718	0		4.32806	1	1	1
11402300	2006	3194.1	0.84	11763.28	53	61.80457	0		-32.2252	1	1	1
11402300	2007	3207.3	0.92	73028	54	58.37621	0		60.72328	1	1	1
11402300	2008	3209.4	0.85	4045	54	58.40344	0		-19.0258	1	1	1
11402300	2009	3185.5	0.81	1843	56	58.09135	0	129.8	58.90582	1	1	1
11402300	2010	3221.5	0.61	0	57	57.47323	0	96.64959	-59.1106	1	1	1
11402300	2011	3343.1	0.61	14528	58	57.07576	0	48.34425	29.77052	0	1	0
11402300	2012	3878.8	0.55	10718	59	56.22358	0	48.47396	-47.0182	0	1	0
11402300	2013	4549.5	0.591	6754	60	62.61127	0	134.9	74.51344	0	1	0
11402300	2014	4690.2	0.416	0	61	62.19777	0	100.1331	-42.274	0	1	0
11402700	1982	2815			55	84.36945	86.92718	106		1	0	0
11402700	1983	2780			56	84.17266	78.84892		26.13455	1	0	0
11402700	1984	2633			57	81.57995	0	52	-32.1093	1	0	0
11402700	1985	2657			58	81.93451	0	90	17.95345	1	0	0
11402700	1986	2621			59	81.68638	0	90	-36.6751	1	0	0
11402700	1987	1833			58	73.86252	100	87	55.99545	1	0	0
11402700	1988	1796			61	73.27394	0		25.52032	1	0	0
11402700	1989	2274				78.89182	0	76	-32.791	1	0	0
11402700	1990	2209			60	79.66048	0		30.68497	1	0	0
11402700	1991	2034			60	75.90954	1.966568		-30.9442	1	0	0
11402700	1995	2029			40	77.36816	0	55.3		1	0	0
11402700	1996	2048			41	77.56348	0	130	79.4662	1	0	0
11402700	1997	1397			42	100	0	108.9	46.65617	1	0	0
11402700	1998	2304			43	80.90278	0	92	-161.376	1	0	0
11402700	1999	1860			44	89.24731	0	113	75.3	1	0	0
11402700	2000	2113			45	78.26313	0	116.6	-112.493	1	0	0
11402700	2001	1895			46	87.36148	0	104.8	136.3329	1	0	0
11402700	2003	2608	0.59	0	48	83.15951	0	88.4		1	0	0
11402700	2004	3263	0.58	65362	49	86.53387	0	154.3	32.16532	1	0	0
11402700	2005	3390	0.51	23770	50	86.51622	0	61	-64.1125	1	0	0
11402700	2006	3409	0.67	31122.21	51	86.56497	0	47.4	72.68597	1	0	0
11402700	2007	3686.2	0.57	238669.3	52	87.57528	4.99159	103.7	45.87577	1	0	0
11402700	2008	3710	0.51	64217	52	87.65498	0	27.4	106.9243	1	0	0
11402700	2009	3839	0.31	3200	55	88.53868	0	146.7	-19.26	1	0	0
11402700	2010	3794	0.38	0	56	88.40274	0	80	-122.834	1	0	0
11402700	2011	4013	0.38	473868	57	89.03564	0		-39.1038	1	0	0
11402700	2012	3440	0.31	367042	58	87.2093	0		-3.0989	1	0	0
11403402	1991	937			37	41.34472	95.30416			1	1	1
11403402	1992	904			38	97.63274	42.25664		-60.5001	1	1	1
11403402	1993	1041			39	100.0096	51.8732		-8.44704	1	1	1
11403402	1994	1883			40	56.22411	54.80616		-17.2001	1	1	1
11403402	1995	958			41	100.0418	56.88935		-57.4517	1	1	1
11403402	1996	1505			42	57.78738	79.00332	48.8	55.265	1	1	1
11403402	1997	1539			43	58.2911	77.64782		63.72388	1	1	1
11403402	1998	845			44	100.0828	74.08284		-107.588	0	1	0
11403402	1999	1328			45	64.00603	91.56626		35.1859	0	1	0
11403402	2000	899			46	98.46497	70.41157		34.28411	0	1	0
11403402	2001	958			47	100.0104	88.30898	18.9	-52.7391	0	1	0
11403402	2002	1460	0.6	26744	48	67.30822	93.0137		20.37994	0	1	0
11403402	2003	1438	0.53	13248	49	66.79416	92.90681		78.4697	0	1	0
11403402	2004	946	0.52	0	50	100.0846	84.88372		-75.0722	0	1	0
11403402	2005	1026	0.27	0	51	100.0292	68.22612		-40.2068	0	1	0
11403402	2006	904.7	0.23	5161	52	100	94.1417		92.10135	0	1	0
11403402	2007	1252	0.28	88863	53	100.008	80.03993		-27.2374	0	1	0
11403402	2008	1217.9	0.37	0	53	100	75.96683		61.33228	0	1	0
11403402	2009	1174.9	0.38	32014	55	100	76.84058	125.3	3.102371	0	1	0
11403402	2010	1206.4	0.22	0	56	100	77.12202	94.39999	-74.9439	0	1	0
11403402	2011	1495.3	0.24	172673.3	57	100	53.63472	9.941385	96.91341	0	1	0
11403402	2012	1138.2	0.23	150527.8	58	100	70.46214	1.95	78.58161	0	1	0
11403402	2013	1125.8	0.157	0	59	92.54752	71.23823	122.11	-36.9549	0	1	0
11403402	2014	1089.5	0.187	13721	60	100	79.02708	80.4	-137.194	0	1	0
11407100	2007	1137.4	0.22	7862	29	100	24.28345			0	1	0
11407100	2008	1211.5	0.18	0	30	100	22.79819	127.4	188.1835	0	1	0
11407100	2009	1217.9	0.04	140	31	100	23.09713		-128.894	0	1	0
11407100	2010	1319.9	0.09	0	32	97.13615	30.93416	89.3	-163.988	0	1	0
11407100	2011	1318.9	0.19	26155	33	97.05815	30.88179	50.95188	38.85012	0	1	0
11407100	2012	1321.4	0.14	28899	34	97.06371	30.87634	0	155.5795	0	1	0
11407100	2013	1398.4	0.108	0	35	97.13959	35.89817		54.56421	0	1	0
11407100	2014	1121.2	0.211	3353	36	96.4324	44.77346	104.1475	-121.254	0	1	0
11413000	2013	1860	0.279	3052.35	59	64.22581	54.13979			1	1	1
11413000	2014	1961.2	0.27	32162	60	60.68733	48.74567		-15.5653	1	1	1
11482100	1983	4600			35	43.26087	10.86957			1	0	0
11482100	1984	4770			36	46.85535	10.48218		-19.3558	1	0	0
11482100	1986	5630			38	33.54885	9.236235			1	0	0
11482100	1987	4503			42	35.41639	11.10371		12.96028	1	0	0
11482100	1990	5587			44	42.51297	20.70879			1	0	0
11482100	1991	6034			45	40.33809	8.170368		-107.419	1	0	0
11482100	1992	4877			46	46.93459	11.79004		80.45451	1	0	0
11482100	1993	6384			47	37.73496	9.022556	12.42308	-10.3144	1	0	0
11482100	1994	3047			48	100	0	91.57233	-7.83307	1	0	0
11482100	1995	7917			49	40.38525	0	62.46154	-24.117	1	0	0
11482100	2004	8258	1.52	2928	59	67.52603	0			1	0	0
11482100	2005	8225	0.82	82161.32	60	67.38724	0		-10.7876	1	0	0
11482100	2006	9872.2	0.81	16338.44	61	66.6741	0		4.837383	1	0	0
11482100	2007	9239.7	0.76	154766.6	62	72.78808	6.360596		32.36461	1	0	0
11482100	2008	9453.4	0.68	102816.4	50	72.0069	0		-20.6736	1	0	0
11482100	2009	9649.2	0.64	20260.32	60	71.98524	0		51.33526	1	0	0
11482100	2010	10048.7	0.21	14989	38	70.39418	0	114.0364	-47.0805	1	0	0
11482100	2011	9829.7	0.3	82612	39	71.78957	0	44.79606	61.60633	1	0	0
11482100	2012	9901	0.18	217974.1	40	71.99273	0	52.06996	-60.9523	1	0	0
11482100	2013	10051.5	0.187	6140.82	41	72.41208	0	146.153	40.54168	1	0	0
11482100	2014	10368.2	0.165	85473	42	73.25476	0	105.1626	-72.482	1	0	0
11495100	1995	1709			34	78.36747	0	25		1	1	1
11495100	1996	1880			35	80.31915	0	35	47.73134	1	1	1
11495100	1997	1870			36	80.24064	0		1.322342	1	1	1
11495100	1998	1868			37	80.21948	0		-26.5384	1	1	1
11495100	1999	2094			38	82.33524	6.638013	123	-19.6251	1	1	1
11495100	2000	2072			39	82.17181	6.708494	114	-9.7599	1	1	1
11495100	2001	2117			40	82.52244	6.565895	55	-4.34722	1	1	1
11495100	2002	2159	0.55	23562	41	82.88097	7.132932		-11.2495	1	1	1
11495100	2003	2067	0.62	2509	42	82.09966	7.353652		41.60542	1	1	1
11495100	2004	2110	0.4	5392.55	43	82.46445	7.203792		-0.98119	1	1	1
11495100	2005	1970	0.41	1477	44	89.3401	13.80711		-71.4828	1	1	1
11495100	2006	1976	0.6	8358.38	45	89.37247	17.30769		29.28703	1	1	1
11495100	2007	2002	0.62	125241.6	46	89.51049	17.08292		45.54249	1	1	1
11495100	2008	1982	0.57	0	46	89.40464	17.2553		73.59829	1	1	1
11495100	2009	1977	0.37	0	48	89.63075	17.29894		-51.4165	1	1	1
11495100	2010	2202	0.26	0	49	90.69028	15.62216		-26.6656	1	1	1
11495100	2011	2203	0.38	111437	50	92.73717	15.61507		-17.3922	1	1	1
11495100	2012	2203.2	0.11	227344	51	92.73784	15.61365		82.50526	1	1	1
11495100	2013	2227	0.089	0	52	92.81544	15.44679		25.23722	0	1	0
11495100	2014	2282	0.211	50980	53	92.98861	15.0745		-139.803	0	1	0
11498100	1998	984			43	51.21951	32.4187			1	1	1
11498100	1999	1072			44	55.22388	29.75746		97.25472	1	1	1
11498100	2000	1072			45	55.22388	29.75746		-38.5997	1	1	1
11498100	2001	995			46	51.75879	30.25126		-34.6561	1	1	1
11498100	2002	1271	0.43	6620	47	62.23446	25.33438		12.05758	1	1	1
11498100	2003	1190	0.34	1785	48	51.26051	27.05882		-11.9285	1	1	1
11498100	2004	1129	0.29	0	49	57.4845	31.17803	107	15.83955	1	1	1
11498100	2005	1310	0.23	7332	50	55.72519	36.64122		54.02818	1	1	1
11498100	2006	1456	0.34	17231	51	56.73077	32.96703	60	-63.2122	1	1	1
11498100	2007	965	0.29	73711	52	83.41969	48.70466	50	64.73859	1	1	1
11498100	2008	1050	0.26	0	52	84.7619	46.19048		19.38132	1	1	1
11498100	2009	1588	0.24	11734	54	53.08564	30.54156	120	-112.624	1	1	1
11498100	2010	1753	0.2	8284	45	52.93782	28.40844	82.5	52.88866	1	1	1
11498100	2011	1619	0.16	40072.21	46	62.32242	39.53057	15	-11.8537	1	1	1
11498100	2012	1630	0.33	91254	47	55.21473	41.71779	13.46154	40.49083	1	1	1
11498100	2013	1567	0.379	10362	48	59.02999	42.75686	0	118.2656	1	1	1
11498100	2014	1719	0.374	31353	49	57.82432	38.97615	85	-154.649	0	1	0
11801800	2001	1406			49	85.81792	36.27312			1	1	1
11801800	2002	1632	0.42	6826	50	85.34927	34.00735		5.504258	1	1	1
11801800	2003	1228	0.38	11306	51	96.79967	40.39088		12.4222	1	1	1
11801800	2004	959	0.5	0	52	100.0834	45.1512		-64.7978	1	1	1
11801800	2005	1292	0.46	0	53	100.0542	36.60991		97.72904	1	1	1
11801800	2006	1375.7	0.44	0	54	100	35.38562		-78.6359	1	1	1
11801800	2007	1152.7	0.46	37108	55	100	35.54264		63.38318	1	1	1
11801800	2008	1466.1	0.43	0	55	100	64.49764		-1.98631	1	1	1
11801800	2009	1485.1	0.3	0	57	100.0067	47.11467		29.32452	0	1	0
11801800	2010	1298	0.22	0	58	100	43.29738		18.83815	0	1	0
11801800	2011	1524.4	0.21	0	59	95.40803	49.85568		38.3519	0	1	0
11801800	2012	1576.4	0.18	13288	60	96.90434	48.21111		-115.083	0	1	0
11801800	2013	1574.3	0.213	0	61	97.88477	48.27542		120.1989	0	1	0
11801800	2014	1775.8	0.186	59518	62	99.79164	42.79761		-154.737	0	1	0
11802300	1993	274			41	92.70073	100			1	0	0
11802300	1994	833			42	72.38895	32.77311	102	49.329	1	0	0
11802300	1995	1439			43	87.14385	19.80542		-6.83802	1	0	0
11802300	1996	1589			44	90.24544	17.55821	50	36.43363	1	0	0
11802300	1997	1785			45	89.63586	15.40616	95	12.87458	1	0	0
11802300	1998	1651			46	88.79467	10.05451	104	-34.7722	1	0	0
11802300	1999	1787			47	89.64745	9.289311	85	30.59828	1	0	0
11802300	2000	1760			48	89.5	9.488636	68	-20.793	1	0	0
11802300	2001	1795			48	89.6936	9.24791		7.300203	1	0	0
11802300	2002	1921	0.6	9770	49	90.68194	8.328996		-7.43066	1	0	0
11802300	2003	1921	0.53	7000	50	90.68194	8.328996		65.06544	1	0	0
11802300	2004	1957	0.46	-1463	51	90.85335	7.256004		-81.6626	1	0	0
11802300	2005	1710	0.54	0	52	93.45029	7.836257	53	15.17937	1	0	0
11802300	2006	1263	0.55	2970	54	97.14964	12.5099		45.96368	1	0	0
11802300	2007	1172.6	0.47	46621.02	55	98.37967	12.02456		2.160885	1	0	0
11802300	2008	1293.2	0.49	0	55	72.47139	12.23322	95	19.55289	1	0	0
11802300	2009	1706	0.48	0	57	75.43963	8.264947		22.07103	1	0	0
11802300	2010	1749	0.44	0	58	76.04346	8.061749		-14.3512	1	0	0
11802300	2011	1935.9	0.36	44399	59	77.60732	18.07945		6.411263	1	0	0
11802300	2012	1710.5	0.42	43686	60	78.81905	29.31891		-47.4065	1	0	0
11802300	2013	1718.6	0.306	0	61	78.81416	29.67532	108	105.3188	1	0	0
11802300	2014	1759	0.298	0	62	76.63445	28.99375	81	-156.981	1	0	0
11806000	2006	1223.2	0.1	0	42	53.67887	45.3728			1	1	1
11806000	2007	1183.7	0.13	8733	43	47.62187	42.57836	76.31805	176.8452	1	1	1
11806000	2008	1221.6	0.12	2421	43	49.24689	38.97348	98.12933	-414.859	1	1	1
11806000	2009	1277.5	0.1	0	43	51.46771	41.49511	124.318	140.9868	1	1	1
11806000	2010	1347.1	0.13	3783	46	53.9752	42.16465	89	234.7461	1	1	1
11806000	2011	1491.3	0.19	57106	47	56.02494	47.94474		-99.2923	1	1	1
11806000	2012	1574	0.19	31077	48	57.88437	45.42567		-69.6203	1	1	1
11806000	2013	1750.5	0.186	0	49	59.28592	40.84547	121	-68.5413	0	1	0
11806000	2014	1876	0.24	33397	50	61.75373	38.11301	90	64.78239	0	1	0
11807200	1980	1434			41	93.72385	16.7364			1	0	0
11807200	1981	1697			42	88.50913	14.1426		29.19524	1	0	0
11807200	1982	1387			43	94.82336	15.93367		21.26633	1	0	0
11807200	1983	1366			44	96.01025	16.76427		-58.971	1	0	0
11807200	1984	1576			45	82.1066	15.22843		19.76654	1	0	0

[Clyde Schechter]

I want to test if there is any difference in outcome in pre-treatment years for both treated and untreated farms. If I just use the interaction term EntryTreat as you have suggested, it will fail to test the difference in outcome as I intended because EntryTreat =0, would include treated farms in post-treatment years as well as untreated farms in pre-treatment years as shown in the data below. So, if I correctly understood your suggestion, I am not sure just using the interaction term would serve my purpose. Please let me know if I am missing anything here. More clarification is highly appreciated.

Yes, you are missing something. It is a bit complicated to explain, but the use of "just" the EntryTreat interaction will work properly provided you use firmid fixed effects and include i.year in the model as well, as advised in #17. For a full explanation, see https://www.ipr.northwestern.edu/wor.../Day%204.2.pdf. Again, this is not the classical diff-in-diff analysis that you have in mind--but that analysis is simply not possible in a situation where different firms begin treatment at different points in time. This is the generalized diff-in-diff analysis.

You may, of course, include your other covariates: the code suggested in #17 was just the essential elements of the analysis.

Per your suggestion, I used dataex output and when I cut and paste the data it messed up the format. I have posted below the formatted data for all the variables that I used in my model for you to look at.

Thanks for trying, and for the extra effort to make it readable. But I think you do not understand the purpose of -dataex-. The use of -dataex- is not intended to create a display of the data that human eyes will easily read. It is intended to create code that can be copy/pasted into Stata's do-file editor and then run to re-create the data set in Stata. The output of -dataex- is usually not readily understood by human eyes; but that is not the purpose. What is usually needed to provide help is a data set that can be run in Stata, and -dataex- provides the way to create that with assurance that it is faithful in relevant ways to your original data. You might ask why we don't just encourage people to attach their Stata data sets. There are two reasons. One is that we usually don't need to see your whole data: just enough of it to run the analysis you are interested in. The other reason is that some of us will not take the risks inherent in downloading anything from a stranger.

[Sunil Dhoubhadel]

Clyde,
Thanks for your suggestion. I will take a look at the reference you suggested.

[Sunil Dhoubhadel]

Clyde,

The link you provided is not working. Could you please provide the title of the paper and the author name?

Thanks
Sunil

[Clyde Schechter]

Hmm! The original link was to a set of slides, not a paper, and I cannot seem to locate it now. Instead of that, have a look at https://www.annualreviews.org/doi/pd...-040617-013507.

[Carlo Lazzaro]

Clyde:
thanks for the reference.