NSC Processing Analysis f

P

paulclarke1

Registered Users (C)
Hi all,

In order to fill in the long cold days in I-485 limbo, I did some analysis on 2001 RD data contained in Rupnet. This is contained in the attached spreadsheet What I was looking for was to see by month the mean, median, mode, mix, max and std dev of calendar days from RD to AD for those lucky people who have been approved with 2001 RD's.

Contrary to what we might percieve NSC seems to have been very consistent over recent months with the average processing time for AD's in Rupnet staying around 400 days. The shortest recorded wait was 314 days (Apr 2001 RD), the longest 562 (Feb 2001 RD).

Rupnet came on line (I think) about May 2001, so my guess is that the early data is skewed to people who were not approved in the NSC's first pass through their month.

If someone with a higher level of statistical knowledge than I could calculate the current processing speed and capacity of the service center, then perhaps we could make a reasonable forecast (based on the sample data available), about future processing times, given what seems to be a significant reduction in the number of applications filed since October 2001.
 
Original Attachment Was Too Big

Sorry, my original attachment was too big. Here is just the summary page.
 
Awesome man...

Give me a Virtual high 5...
good job..but ..I dunno abt Sept - I am a sept filer,and its 410 days already for me...so I dunno man..
have a great weekend everyone
Cheers
Jack
 
Good Luck Jack

I don't know either about September, with only 9% of the RD's approved it is a very small sample size. I suspect however that you are close. At least closer than I am at 228 days (989 since I entered the process).
 
Very Good analysis Paul, but one catch. The analysis ignores the cases that are yet pending. While your analysis gives a good picture of time frames for cases approved, it does not take into account the fact there may be cases beyond these time frames that have not been approved. Of course, we need to eliminate exceptions where people entered initial data but never returned to update approval/RFE details.

Maybe if you include max/min time case pending, it will give even a better picture. Also take into account RFE cases. So we are talking about "Cases acted upon" rather than cases approved.

This can be calculated by the formula (Cases approved + RFEs - RFEs approved) / Filers

IF you go by this formula, you will note that for July, this calculates to > 80%. My take on doing this analysis for previous 3 months is that once NSC hits a "Cases acted upon" ratio of about 75%, it starts acting on cases for next month. This happened when NSC started RD Jul and Aug 2001 cases. They seem to have halted processing now for other reasons, amongst them I believe from reading other boards is that NSC allocated extra adjudicators to meet the the July 2001 rush. That's why you saw a consistent pace for July approvals. Now with the July rush over (being met with the 75% case acted upon threshold), they simply removed the extra adjudicators, thereby bringing approvals to a crawl.

Cheers. :rolleyes:
 
I am Not Sure

T-Rex
I like the inference about when NSC hits 75% of cases acted upon then generally are in a position to move on to the next months processing. That is a great insight and one that had not occured to me. That might be a useful rule of thumb to keep in mind. However I think it is a little early to conclude that the adjudicatation pattern has changed. Remember we are now starting to look at smaller population so we can expect to see a lower absolute number of approvals.

I did not set out to ignore those people who were still pending but I am not sure that you can really draw any conclusions from the data where the AD has not been loaded. There is really no way to know if the record that does not have an AD is still pending or has been approved but no data has been loaded. You would need to survey these people to tell what there status is.

I realize that that is not a satisfying answer. If you look at August you will probably get a feel for the range of likely adjudication times (somewhere between 350 and 500 days), this is consistent with the standard deviation of the data as well. All of this only serves to highlight why averages are often misunderstood. Being above or below average is not a measure of good or bad, it is just an indication of central tendancy.

If you want to do more analysis send me a private message with your email address and I will send you the entire file. I would love to see what you came up with.
 
You must log in or register to reply here.
Top