I postulate some assistance from the increase empiricists out there. If yous aren't i of them, halt reading. Continuing volition be a waste matter of your time.
For researchers studying economical growth, i of the criterion resources for cross-country information has been the Penn World Table. My 1992 newspaper amongst David Romer together with David Weil (my most cited newspaper past times a large margin) used this resource, as convey numerous other papers inward this literature. In my intermediate macro book, I acquaint a twosome of figures presenting some of these data.
Here's the problem: It seems that the information convey changed substantially in the most recent revision, together with I cannot figure out why.
My intermediate macro text shows a scatterplot of per capita income together with the investment portion of GDP. These 2 variables are strongly positively correlated. When revising this figure amongst the newest data, I flora that the correlation declines substantially (though is yet positive). When I looked into the beginning of the change, I found that the historical estimates of the investment portion of gross domestic product convey changed, inward some some cases past times a lot.
Let me order yous an example. Take the investment share for Republic of Ghana inward the twelvemonth 2000. According to version 6.2 of the data, the investment share was about five percent. In version 7.0, it was close 21 percent. This is one of the larger changes I convey found, precisely it is not the alone terra firma for which at that spot are sizable changes inward the reported investment portion of GDP.
I empathize that the changes may be related to novel information close the relative cost of investment goods. But the changes seem also large to live on explained together with thus easily, although mayhap I am incorrect close this. If anyone tin shed calorie-free on the matter, I would live on greatly appreciative. Send me an e-mail if yous tin help.
-----
Update: I convey non yet fully figured this out, precisely readers convey sent me some useful links. If yous are interested, click here, here, together with here. Sumber https://gregmankiw.blogspot.com/
For researchers studying economical growth, i of the criterion resources for cross-country information has been the Penn World Table. My 1992 newspaper amongst David Romer together with David Weil (my most cited newspaper past times a large margin) used this resource, as convey numerous other papers inward this literature. In my intermediate macro book, I acquaint a twosome of figures presenting some of these data.
Here's the problem: It seems that the information convey changed substantially in the most recent revision, together with I cannot figure out why.
My intermediate macro text shows a scatterplot of per capita income together with the investment portion of GDP. These 2 variables are strongly positively correlated. When revising this figure amongst the newest data, I flora that the correlation declines substantially (though is yet positive). When I looked into the beginning of the change, I found that the historical estimates of the investment portion of gross domestic product convey changed, inward some some cases past times a lot.
Let me order yous an example. Take the investment share for Republic of Ghana inward the twelvemonth 2000. According to version 6.2 of the data, the investment share was about five percent. In version 7.0, it was close 21 percent. This is one of the larger changes I convey found, precisely it is not the alone terra firma for which at that spot are sizable changes inward the reported investment portion of GDP.
I empathize that the changes may be related to novel information close the relative cost of investment goods. But the changes seem also large to live on explained together with thus easily, although mayhap I am incorrect close this. If anyone tin shed calorie-free on the matter, I would live on greatly appreciative. Send me an e-mail if yous tin help.
-----
Update: I convey non yet fully figured this out, precisely readers convey sent me some useful links. If yous are interested, click here, here, together with here. Sumber https://gregmankiw.blogspot.com/
