neilalex / ad-replication Goto Github PK

1. In line 84-89 of Parametrs.do @neilalex you have some analysis on inflation which is not transparent enough.

• It would be helpful to make more clear how you get the PCEI indices form BEA. The website you posted in your comments helped me only to find monthly PCE of 2016. How you computed the annual PCEI?

• What is the formula you use for the "Hand-tuned inflation calibration factors"?

2. In line 101-114 you have some analysis on top-coded wages.

• What is the usage of the top wages?
• It is not clear how you get the annual to wage i.e. what you call "topWageYYYY". IPUMS-CPS website you have posted has only period-specific variables, which is not clear which one is the topWageYYYY.
• What the "topCodeWageFactor" does and how you compute it?
• I suspect that the value of the "unknownWage" is an assumption. Could you elaborate on what this implies and how it affects the analysis?

1. In this line it seems to me that it should be table7_k'_index' . I do not understand what the following line does : naxSaveOutput "Table6 7 index=index' k=k'" "rtiShare".

2. It seems that you have skipped the wage analysis. That is, Panel B of table 7.

The variable "serviceShare" in TableAppendix1 was weighted not for all the occupations present in the database, but only for those with no college education. So, it would be a portion of them and this is correct for TableAppendix1.

The problem arises when this measurement is also used in Figure 6 for the regression. and I'm not sure if it's correct to use this measurement.

I'm not sure if this is the right way to identify this variable or if it's better to weigh the no-college individuals in the "serviceShare" against all individuals for figure6.

If someone could clarify this for me, I would be grateful.

In etlCz.do I see you rename cntygp97 and cntygp98. I think in the IPUMS data I have cty_grp70 and ctyprp1980 -- is there a reason for the difference?

PS_1: I think that I was misled from the list of variables you suggest to download. You have cntygrp97 while you should have cntygp97.

PS_2: I was also confused because here you use cntygp* while you propably want to rename cty_grp70 and ctygrp1980 that you find in Dorn's crosswalks. Initially, I thought that this was due to changes in IPUMS data but It is clear now that it is due to changes in Dorns' crosswalks.

1. In the file czRtiShares.do @neilalex mentions RIOS. It is not clear what RIOS stands for.

2. @neilalex you create an RTI by industry-state-cz (ISC). However, it is not clear why an RTI by CZ would not be enough and how you manage to have an RTI by ISC; provided that a CZ might belongs to more that one US states.

In etlIpumsCzOcc.do @neilalex in line 31 divides by 10. It is not clear to me why you did that.

For 2005 the OCC (provided by IPUMS USA: ACS ) is a 3-digit code?! Dorn's panel of occupations allocates occupations (OCC) to his occ1990dd codes. So why you changed OCC dividing by 10?

Dorn has changed [E2] and [E3] crosswalks in his website. You should change the Macros adding "_corr" in the end of the crosswalks' names (this applies only for 1970 and 1980).

In the folder Utilities @neilalex creates several routines. It is not clear to me what is he after i.e. what was missing in the standard STATA built-in commands. Please elaborate more on what each of your Utilities does.

Also, It is not clear to me when you call these Utilities what task they perform and what is the syntax rules you follow. For instance, what naxMerge "m:1 year occ using occ1990ddAssembled" 1 0 1 "etlIpumsCzOcc" does, and how it differs from naxMerge "1:1 czone ind1990 using temp/cz_industry_exRti, keepusing(rtiShare)" 0 0 1 "czIV"?