Understanding Society User Support

Hello,
Please could you let me know if the following way is valid for merging household data into individual files. I am asking because I am aware that household identifier changes across waves if the household composition changes.

For my analysis, I want to have a measure of household income at Waves 1 through to 5. Therefore in order to do this, I brought the household identifiers (w_hidp) from xwaveid.dta for Waves 1 to 5, into the e_indresp.dta files using a m:1 merge. I then matched the household income from a_hhresp.dta to e_indresp.dta directly using a m:1 merge using a_hidp that I had previously brought in from the xwaveid.dta file. I repeated this for Waves 2 to 5. Is it correct to use the household identifier from the xwaveid.dta file, in this way? I thought it was, because even if the individual changes household over time, matching income from household files on the household id from a specific wave ensures that the observation is correct for that point in time.

The alternative method would be to merge household income into individual level files for each wave and then perform a 1:1 merge from individual files from Waves 1, 2, 3 and 4 into the Wave 5 file, matching on pidp. However, when I used this method, it resulted in 4,000 more cases of missing data for Wave 1.

Any advice that you can give on this issue would be much appreciated.

Many thanks,
Dharmi