Is the US Cooking the Books on GDP?

An old school trick economists use to measure the true economic growth of countries where their governments are not...um...known for publishing accurate and true figures

COUGH COUGH
ARGENTINA

COUGH COUGH
CHINA

WHEEZE WHEEZE
VENEZUELA

is instead of taking the liar's word for it, they instead substitute electrical consumption.  This is a reasonably good proxy, unaffected by inflation or distortions from government bureaucrats and can be easily measured.

So I decided to do the same thing with the United States.

Here is electrical consumption vs. Real GDP going back to 1949.

Because of the varying axes it is important to index both to 100 in 1949 to see whether one is growing faster than the other.  Here is that same chart indexed:

If you eyeball it, it seems the opposite of what I was suspecting was happening.  Energy consumption is growing faster than GDP suggesting we've been UNDERSTANDING our economic growth.  I double checked the figures , thinking I had erred, but they prove to be the correct.

Understand, however, I still actually think this is an error because if anything the US has become more efficient with its energy and I have seen GDP/unit of energy figures proving this.  But here's the interesting thing.  Even with this error (that goes against my original case) this data series actually ends up proving it even more so, notably by what I spy with my little eye in the green circle.

Since the chart is indexed you don't really see it, but starting in 2002 energy consumption flatlined while GDP continued to grow.  Matter of fact, GDP has grown by 14% since 2002 while energy consumption has grown only by 1%.

This is what I feared was happened.  Where the government says,

"Yep!  you're all 14% richer now!"

When the energy consumption does not prove so.

Now it is hard to say whether this is because of measuring differences as the chart shows the two do not perfectly correlate with one another, but NEVER in the history of the two data sets has GDP been a full 13% higher than what electrical consumption would suggest.

I can only want to see how much larger the differences will be once we adapt the "New GDP" measure.

(also, any thoughts on the discrepancy or data source from the EIA would be appreciated).