The Impact of Music Downloads and P2P File-Sharing: Appendix 4

Appendix 4: Determinants of CD album purchases in the Canadian population for the year 2005

Dependent variables	Number of CD albums			Log Number of CD albums			Square root no of CD albums
Estimation model	OLS			OLS			OLS
Independent variables	b	t-value	sig	b	t-value	sig	b	t-value	sig
Price of CDs	-0.036	-0.36		-0.002	-0.30		-0.004	-0.34
Number of P2P	-0.807	-1.31		-0.017	-0.80		-0.056	-1.11
Number of CDs ripped	2.941	3.24	***	0.136	3.99	***	0.299	3.70	***
Number of promotional	0.115	0.20		-0.008	-0.28		-0.004	-0.07
Number from private web	2.087	1.74	*	0.078	1.86	*	0.192	1.85	*
Number of copied MP3s	0.323	0.45		-0.007	-0.24		0.007	0.12
Number of purchased MP3s	0.066	0.11		0.026	0.84		0.033	0.52
Number of DVDs	0.866	2.15	**	0.071	3.41	***	0.125	2.93	***
Number of videogames	2.371	2.86	***	0.112	3.61	***	0.245	3.43	***
Number of cinema tickets	0.092	0.24		0.037	1.54		0.044	1.00
Number of concert tickets	3.158	3.38	***	0.148	4.52	***	0.322	4.09	***
Income 10 to 20	-0.155	-0.10		-0.004	-0.03		-0.021	-0.09
Income 20 to 40	-0.158	-0.13		0.056	0.46		0.041	0.22
Income 40 to 60	1.425	0.96		0.162	1.30		0.240	1.19
Income 60 plus	1.345	1.07		0.183	1.53		0.255	1.38
Quality increase	-0.350	-0.36		0.018	0.33		-0.006	-0.06
Quality decrease	1.933	1.31		0.100	1.42		0.202	1.39
Interest very strong	4.370	3.10	***	0.463	3.28	***	0.730	3.47	***
Interest somewhat strong	3.118	2.45	**	0.422	3.03	***	0.614	3.03	***
Interest moderate	-0.003	0.00		0.135	0.98		0.136	0.70
Interest somewhat low	-1.203	-0.91		-0.018	-0.11		-0.093	-0.41
Internet: very skilled	-2.930	-1.38		-0.161	-1.41		-0.335	-1.46
Internet: skilled	-5.230	-2.89	***	-0.217	-2.12	**	-0.517	-2.60	***
Internet: somewhat skilled	-3.387	-1.97	**	-0.110	-1.09		-0.307	-1.59
Internet: not very skilled	-4.790	-2.75	***	-0.221	-2.08	**	-0.499	-2.52	**
Age 15 to 19	1.735	1.06		0.055	0.44		0.159	0.76
Age 20 to 24	2.144	1.36		0.135	1.10		0.273	1.32
Age 25 to 34	3.237	2.12	**	0.132	1.09		0.325	1.62
Age 35 to 44	4.602	2.26	**	0.246	1.97	**	0.511	2.25	**
Age 45 to 54	4.048	2.72	***	0.247	2.13	**	0.486	2.50	**
Age 55 to 64	4.284	2.67	***	0.249	2.12	**	0.495	2.48	**
Gender (men=0, women=1)	-0.555	-0.83		-0.043	-0.93		-0.082	-0.99
Region (Quebec=0, rest=1)	-1.632	-2.14	**	-0.115	-2.56	**	-0.211	-2.48	**
Constant	4.951	1.92	*	1.500	6.50	***	1.979	5.38	***
Number of observations		1,387			1,387			1,387
Population size	15,291,433			15,291,433			15,291,433
F-value		5.42	***		7.63	***		6.94	***
R-square		0.25			0.24			0.26

* p<0.10; ** p<0.05; *** p<0.01. All regressions are estimated with a constant. The number of observations is smaller than the total number of respondents in the survey due to missing values; e.g. Price of CDs was answered by 1,575 survey participants.
Source: own calculations based on Decima Research survey.

Appendix 4: Determinants of CD album purchases in the Canadian population for the year 2005 (continued)

Dependent variables	Number of CD albums			Number of CD albums
Estimation model	Poisson			Negative binomial
Independent variables	M.E.	t-value	sig	M.E.	t-value	sig
Price of CDs	-0.018	-0.19		-0.015	-0.20
Number of P2P	-0.345	-1.30		-0.150	-0.70
Number of CDs ripped	1.828	4.17	***	1.537	4.08	***
Number of promotional	-0.008	-0.03		0.003	0.01
Number from private web	0.999	2.07	**	0.807	1.81	*
Number of copied MP3s	-0.021	-0.06		0.064	0.22
Number of purchased MP3s	0.021	0.06		0.111	0.37
Number of DVDs	0.692	2.46	***	0.745	3.20	***
Number of videogames	1.515	4.24	***	1.296	3.93	***
Number of cinema tickets	0.147	0.52		0.346	1.37
Number of concert tickets	2.058	4.48	***	1.610	4.40	***
Income 10 to 20	-0.249	-0.15		-0.402	-0.24
Income 20 to 40	-0.267	-0.20		-0.197	-0.15
Income 40 to 60	1.591	1.07		1.596	1.05
Income 60 plus	1.332	1.03		1.448	1.09
Quality increase	-0.271	-0.41		-0.352	-0.56
Quality decrease	1.660	1.75	*	0.975	1.22
Interest very strong	6.817	3.44	***	7.242	3.80	***
Interest somewhat strong	5.757	2.97	***	6.071	3.24	***
Interest moderate	1.745	1.00		2.134	1.24
Interest somewhat low	-0.459	-0.25		-0.116	-0.06
Internet: very skilled	-2.748	-1.85	*	-2.200	-1.62
Internet: skilled	-4.178	-3.26	***	-3.302	-2.93	***
Internet: somewhat skilled	-2.831	-2.19	**	-2.149	-1.87	*
Internet: not very skilled	-3.873	-3.14	***	-3.111	-2.79	***
Age 15 to 19	3.551	1.96	*	1.544	1.03
Age 20 to 24	3.816	2.12	**	1.873	1.22
Age 25 to 34	4.269	2.46	**	1.949	1.32
Age 35 to 44	6.264	3.08	***	3.813	2.36	**
Age 45 to 54	5.730	3.22	***	3.417	2.31	**
Age 55 to 64	6.153	3.10	***	3.372	2.15	**
Gender (men=0, women=1)	-0.501	-0.84		-0.522	-0.97
Region (Quebec=0, rest=1)	-1.449	-2.19	**	-1.742	-2.99	***
Constant	1.461	4.90	***	1.559	5.75	***
Number of observations		1,387			1,387
Population size	15,291,433			15,291,433
F-value		9.22	***		8.98	***
R-square		-			-

* p<0.10; ** p<0.05; *** p<0.01. All regressions are estimated with a constant. Marginal effects (M.E.) are given with the t statistics of the underlying coefficients. M.E. are calculated at the means of the regressors in the case of continuous data and for discrete change from 0 to 1 in the case of dichotomous variables. The number of observations is smaller than the total number of respondents in the survey due to missing values; e.g. Price of CDs was answered by 1,575 survey participants.
Source: own calculations based on Decima Research survey.