I'll use the US as an example. I have three data series
- nominal GDP $(Y)$
- real GDP in 2005 USD $(\bar{Y})$
- the GDP deflator $(d)$, with 2005 as the base year, so $d_{2005} = 100$
I want to change the base year to 2000. Are these calculations accurate? I use the notation $\$_{t}$ for USD in year $t$ prices to help myself keep the units straight.
My goal is $\bar{Y}_{t} \ \$_{2000}$. \begin{align} \frac{d_{2000}}{d_{t}} \cdot Y_{t} \ \$_t &= \frac{Y_{2000} \ \$_{2000}}{\bar{Y}_{2000} \ \$_{2005}} \cdot \frac{\bar{Y}_{t} \ \$_{2005}}{Y_{t} \ \$_{t}} \cdot Y_{t} \ \$_t \\ &= \frac{Y_{2000} \ \$_{2000}}{\bar{Y}_{2000}} \cdot \bar{Y}_{t} \\ &= \bar{Y}_{t} \ \$_{2000} \cdot \frac{Y_{2000}}{\bar{Y}_{2000}} \end{align}
I think those are all the correct unit cancellations, but now I'm stuck with the unitless quantity $\frac{Y_{2000}}{\bar{Y}_{2000}}$, so I don't know how to complete the conversion.
Am I doing this right?