0
$\begingroup$

I'm trying to collect current GDP data for India from 1990-2016. But, India changed its base year in 2011. So, to get a continuous series, do I have to change the data to a common base?

$\endgroup$
  • 1
    $\begingroup$ Possible duplicate of Why do we change the base year? $\endgroup$ – luchonacho Jun 6 '17 at 9:43
  • $\begingroup$ For the lazy one (or if you want to have a benchmark to compare your calculation), see here. $\endgroup$ – luchonacho Jul 6 '17 at 11:13
  • $\begingroup$ in some cases that base year is not mentioned in data set or by data provider, how can we find out which year is the base? $\endgroup$ – bio tech Dec 7 '18 at 7:00
1
$\begingroup$

Yes,

If you initially had GDP data at some base year $Year\;x$ for some $Year\;y$

$$I_{\frac{Year\;y}{Year\;x}} = \frac{GDP_{Year\;y}}{GDP_{Year\;x}} \times 100$$

If your base year has now changed to some base year $Year\;a$ for some $Year\;b$ , $a > x$, then you should convert the GDP data from year $b$ onwards to reflect the initial base year, $x$.
$$I_{\frac{Year\;b}{Year\;x}} = I_{\frac{Year\;b}{Year\;a}} \times I_{\frac{Year\;a}{Year\;x}} $$

Further proof, $$\frac{GDP_{Year\;b}}{GDP_{Year\;x}} = \frac{GDP_{Year\;b}}{GDP_{Year\;a}} \times \frac{GDP_{Year\;a}}{GDP_{Year\;x}} $$

This way you will be able to get a continuous and consistent series based on some base year, $x$. If there were multiple base years, the data would be meaningless since it would reflect on changes with regard to different periods of time - ie, you would not be able to draw a definite conclusion.

An index compares the value of the current year against the value in the base year. In this case, you might be looking for the current GDP value for every year from your data-set. As an example, if $$I_{\frac{2001}{2000}} = 110, \; and \; GDP_{2000} = 200, \; then \; GDP_{2001} =220.$$
When comparing data with changing base years, for eg, $$I_{\frac{2001}{2000}} = 110, \;I_{\frac{2002}{2001}} = 105,$$ You can see that this does not mean the GDP in 2002 is $10\%+5\%$ higher than 2000. Instead, the actual GDP =

$$I_{\frac{2002}{2000}} \times GDP_{2000} = I_{\frac{2002}{2001}}\times I_{\frac{2001}{2000}} \times GDP_{2000} = 115.5 \times GDP_{2000} = 231$$ or you can break this down to $$GDP_{2000} = 200$$ $$GDP_{2001} = GDP_{2000} \times I_{\frac{2001}{2000}} = 220$$ $$GDP_{2002} = GDP_{2001} \times I_{\frac{2002}{2001}} = 231 $$ Thus this is why you have to change your indexes to reflect on the same base year, since your calculated values of actual prices will differ.

$\endgroup$
  • $\begingroup$ Thanks Joshua. Just a quick question. Isn't current data supposed to be based on current prices? That's what got me confused. Why is current data affected by base year changes? $\endgroup$ – Chirag Yadav Jun 6 '17 at 9:18
  • $\begingroup$ @ChiragYadav, I have edited my answer to include an example. Hopefully this will help. The index is the ratio of current prices of the current year, against the prices in the base year. Thus when the base year changes, the base year price changes too. Therefore indexes with different base years cannot be compared with each other, since a 10%(Index 110) growth from 2000 to 2001 and a 5%(Index 105) growth from 2001 to 2002, does not equal a 10+5% growth in 2000 to 2002, as shown in my example. Your continuous series can be charted by taking the indexes of 90'-16' against 90', or the actual GDPs. $\endgroup$ – Joshua Chin Jun 6 '17 at 10:01
  • $\begingroup$ Also, here's a question similar to yours that i just found : How to handle change in base year of index? $\endgroup$ – Joshua Chin Jun 6 '17 at 10:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.