Jean-Thomas Bernard, Nadhem Idoudi, Lynda Khalaf, and Clément Yélou,
"Finite Sample Inference Methods for Dynamic Energy Demand Models",
Journal of Applied Econometrics, Vol. 22, No. 7, 2007, pp. 1211-1226.

The data are annual data for the Province of Québec (Canada), over the
period 1963-2000, for six sectors of energy use: (i) residential, (ii)
commercial, and (iii) four manufacturing sectors [paper and allied products,
primary metals, refined petroleum and coal products, and all other
manufacturing industries]. We provide a separate .txt data file for each
sector.

The six .txt files are ASCII files in DOS format. They are zipped in the
file biky-data.zip. Unix users should use "unzip -a".

Sector specific total energy demand (measured in terajoules) is modelled as
a function of its own lag and the real energy price, real income, and a
measure of temperature (heating degree days).  Temperature effects are
excluded from the demand equation for the four manufacturing industries. In
the residential sector, energy demand and real disposable income are
measured on per household basis. 

Data on expenditures and quantities by sector are from Statistics Canada. In
particular, for the manufacturing industry, we use the annual Statistics
Canada publication on consumption of purchased fuel and electricity in
the manufacturing industries, catalogue 57-208 (which pertains to the annual
census on the manufacturing sector). Publication of this catalogue was
discontinued after 1984. Some variables were still available from Statistics
Canada (specifically, total energy expenditures for 1985, 1986, and 1990, and
quantities of electricity for 1997-2000). Thus, we have completed the
missing data [quantities and prices (1985-1990), quantities (1997-2000)]
using other data sources. For more information, see Nadhem Idoudi, "La
demande énergétique du secteur manufacturier québécois : analyse de la
stabilité structurelle", Master's Thesis, 2003, Université Laval.

The data file for each of the four manufacturing sectors consists of 6
columns that correspond to the following variables:

Column 1: years
column 2: Total energy demand
column 3: constant (ones)
column 4: Lagged total energy demand
Column 5: Real energy Price
Column 6: Real sector industrial Gross Domestic Product (GDP)

For the residential and commercial sectors, there are 8 columns as follows: 

Column 1: years
column 2: Total energy demand
column 3: constant (ones)
column 4: Lagged total energy demand
Column 5: Real Energy Price
Column 6: Real disposable income (for Residential sector)
Column 6: Real commercial GDP (for Commercial sector)
Column 7: Heating Degree Days
Column 8: Lagged Heating Degree Days.