From The Penn World Tables (PWT7):
Before 1970, the price levels of the three GDP components had been going in parallel. In the late 60s, the price level of government consumption went up (the Nixon Shock + Vietnam War?), and then it steadily increased since after 1980. Investments also behaved differently since then.
Is this price behavior due to the budget deficit (for government consumption) and trade deficit (for investment deflation) or there're some other important factors?
The price level of government consumption (education, parts of health care, police etc.) consists mainly of wages.
Since the productivity (as we measure it) of government production often does not increase too much, the price level increases about as much as the level of wages. This, contrary to the price level of consumer expenditure which is driven down by productivity, increases in manufacturing as well as cheap imports.
The price level of investment goods might have gone down because of 'hedonic'calculation, i.e. statisticians calculate the dollar price of one unit of computing power instead of the price of one computer. In the new national accounts, military equipment is however also (and rightly so, IMO, as many military items have a clear second hand value and a well developed second hand market exists) considered to be an investment good, this gear may have another pattern of development than computers.
In a general sense, I suspect that the investment price level shown is only valid for machinery and the like. Most investment is however in land related items like dwellings, roads, bridges and other buildings as well as the prices of these and of the land itself, have not gone down that much (to the contrary, even, though there are quite some differences between the pre- and post-2008 period).
Here is some useful background information for the EU which shows that price levels of investment goods (broad definition) can differ up to 400% between countries.
