Ana Maria Herrera and Sandeep Kumar Rangaraju, "The Effect of Oil
Supply Shocks on U.S. Economic Activity: What Have We Learned?",
Journal of Applied Econometrics, Vol. 35, No. 2, 2020, pp. 141-159.

All data files are ASCII files in DOS format. They are zipped in the
file hr-data.zip. Unix/Linux users should use "unzip -a".

The variables are organized in columns. The first row lists the
variable name. The first column in each file provides the dates of the
observation. All files are comma-separated value (csv) files.

The following files contain the time series data used in the article

*****

K09.csv: monthly data (1973:2-2016:12)- Total number of observations: 527,

1. World Oil Production: Percentage change in global crude oil
production (expressed in logs)

2. Real economic activity: Kilian's Global Real Economic Activity
Index,

3. Oil Price: Log of real oil price		

*****

KM12.csv: monthly data (1973:2-2016:12)-Total number of observations: 527, 

1. World Oil Production: Percentage change in global crude oil
production (expressed in logs)

2. Real economic activity: Kilian's Global Real Economic Activity
Index,

3. Oil Price: Log of real oil price.

*****

KM14.csv: monthly data(1973:2-2016:12)-Total number of observations: 527,

1. World Oil Production: Percentage change in global crude oil
production (expressed in logs)

2. Real Economic Activity: Kilian's Global Real Economic Activity
Index,

3. Real Oil Price: Log of real oil price

4. Inventories: First difference of inventories

*****

BH18.csv: monthly data (1958:1-2016:12)-Total number of observations:708, 

1. global oil production: Global crude oil Production (million
barrels/day)

2. real WTI spot price: Real WTI spot oil price (deflated by US CPI),

3. OECD+6NME IP: World IP

4. delta inventories: 100*change in inventories as fraction of last
period's oil production

5. real RAC: Real Refiners Acquisition cost of crude oil imports
(starts in 1974M1 before that 1 indicates missing data)

6. KilianREA: Kilian's Global Real Economic Activity Index

*****

LN12.csv: monthly data(1973:1-2016:12)-Total number of observations: 528, 

1. poil_index: Crude Oil (petroleum), Price index, 2005 = 100, simple
average of three spot prices; Dated Brent, West Texas Intermediate,
and the Dubai Fateh

2.poil: Crude Oil (petroleum), simple average of three spot prices;
Dated Brent, West Texas Intermediate, and the Dubai Fateh, US$ per
barrel 

3. q_oil:Global Oil Production in barrels per day

4. wti: oil price (WTI quality)

5. cpi_us: US CPI Seasonally adjusted

6. y_us: US Industrial Production

7. us_ppi:US PPI (all commodities) level

8. row_imp: Rest of World Imports 

9. worldip: World Industrial Production		
	
The VAR uses five variables in the benchmark system:

1. Log of oil production

2. Log of real spot oil price

3. Log of US Industrial Production

4. Log of rest of the world imports

5. Log of U.S. producer price index.

*****

BP13.csv: quarterly data(1947:q2-2016:q4)-Total number of
observations: 279, 

1. q_oil: Percentage change in global crude oil production (expressed
in logs)

2. p_oil: Percentage change in real oil price (expressed in logs) 

3. GDP:Percentage change in GDP (expressed in logs)

4: CPI: Percentage change in US CPI (expressed in logs)


*****

Data sources:

World oil production is obtained from the Monthly Energy Review
published by the Energy Information Agency (EIA) and is measured in
thousands of barrels per day. Kilian's(2009) index of global economic
activity is computed as the cumulative average rate of increase in
bulk dry cargo ocean freight rates, deflated by the U.S. CPI and
linearly detrended. The log of the real oil price is measured as the
log difference between the refiners acquisition cost (RAC) of imported
crude oil and the U.S. CPI. The refiners' acquisition cost is provided
by the EIA starting in January of 1974. We extrapolate the data from
January 1974 until January 1973 following Barsky and Kilian (2002).

Data for the CPI are obtained from the St. Louis Fed FRED database. 
The updated time series for this index is available from Lutz Kilian's
website at http://www- personal.umich.edu/~lkilian/reaupdate.txt. See
Kilian (2009) and Kilian and Zhou (2018a) for a detailed description
of the methodology employed to compute the index and a comparison with
alternative indicators of global real economic activity.

RoW output for LN12 model is calculated as global exports to the world
net of global imports from the U.S. and from oil exporting countries.
These data are obtained from the IMF. RoW output is expressed in real
terms and measured in logarithms. Because data for inventories are not
available for countries other than the US, we follow KM14 in using
data from the EIA for total US crude oil inventories, as well as US
and OECD petroleum stocks to compute a proxy. The inventory proxy is
calculated by multiplying US crude oil inventories by the ratio of
OECD to US petroleum stocks.  
		
Further details about estimation and the data can be found in the
paper