Analysis of Express Entry Statistics

[picklee]

I am posting this initial analysis of the 2015 Express Entry (EE) report by IRCC. There are some good statistics in the report that support an analysis of EE trends, rates of acceptance, and issued invitations to apply (ITA).

Here is the report:
http://www.cic.gc.ca/english/resources/reports/ee-year-end-2015.asp

TLDR; rates of ITAs from 2015 and 2017 targets for FSW and CEC programs suggest about 57k additional ITAs will be issued this year, about 2,900 over the next 20 draws.

Most data in the report were year-end totals for 2015, as of 03JAN2016.

Other attempts have been made to analyze trends in ITAs based on target program quotas. The problem with these analyses, as pointed out previously, is that each ITA may have more than one applicant associated with it. Moreover, IRCC needs to take into consideration the number of ITAs that will expire or be declined as well as the number of applications that will be rejected.

ANALYSIS:

Here are some summary statistics and calculations:

31,063 ITAs issued
-21,562 Applications submitted
=9,501 (30.6%) ITAs declined/expired

37,424 Total applicants
=1.7356 Applicants per application
-15,246 Applicants under review
=22,178 Applicants processed
=12,778 ITAs processed

14,058 (63.4%) Applicants approved
=0.91 ITAs per approved applicant

These results support the fact that IRCC needs to issue more than one ITA for every approved PR, despite the fact that each ITA is associated with more than one applicant, on average.

As mentioned in the IRCC report, they were still processing 15k of 2015 applicants as of 03JAN2016. Based on the rates above, about 10k of those 2015 applicants would have been approved in 2016 or about 25.8% of total 2015 applicants/applications.

2015 admissions from EE:
13.5% FSW + CEC
1.1% PNP

http://www.cic.gc.ca/english/resources/publications/annual-report-2016/index.asp#s1.1

The majority of applications for PR in all these classes were not submitted by EE, probably because it was the first year for EE and many paper applications were processed.

2015 ITAs by program:
78.6% FSW + CEC
13.2% PNP

OUTLOOK:

We can expect EE application rates to increase, but we can probably expect the other rates to stay about the same: 0.91 ITAs per approved applicant; 63% applicants approved; 26% carryover from previous year.

33,782 ITAs were issued in 2016. Assuming the same average number of applicants per ITA as in 2015, that represents approximately 58,633 total applicants. We can expect 15,144 (25.8%) of those applicants to be approved for PR in 2017. Those count towards 2017 targets for the economic and PNP programs.

2017 program targets:
73,700 FSW + CEC (range 69,600-77,300)
51,000 PNP (range 49,000-54,000)

http://www.cic.gc.ca/english/department/media/notices/2016-10-31.asp

2016 applicants approved in 2017:
11,903 FSW + CEC
1,999 PNP

If we assume that rates stay the same and that 100% of FSW and CEC applications in 2017 come through EE, then IRCC would need to issue approximately 67,909 ITAs to meet the 2017 targets for those two programs.

13,388 ITAs have already been issued to-date in 2017. Assuming, 78.6% of those are for FSW and CEC like in 2015, then IRCC would need to issue approximately 57,386 ITAs to meet the 2017 targets for FSW and CEC.

http://www.cic.gc.ca/english/express-entry/past-rounds.asp

That works out to about 2,900 ITAs over the next 20 draws.

This figure may be optimistic, given that the percentage of PR applications that come through EE may be less than 100%. Also, IRCC has settled fewer than the target number in the CEC class in the past.

 

[Saskatchewan_Regina]

Outdated, good effort though

 

[picklee]

Outdated, good effort though

Any reason to expect that number of ITAs per accepted applicant have changed since 2015? Number of applicants per ITA? Approval rates? I am using invitation data from 2016 and 2017 as well. We are talking about a sample size of n=31,063.

We know that CRS point allocation has changed slightly with job offers going from 600 to 50-200 points and international students can earn up to 30 points for Canadian degree. I don't see that drastically affecting acceptance rates.

 

[astralsource]

+1 good work

 

[robb83]

Carrying on stat conversation here

Unfortunately I cannot get any indication of how they decide a cut off point, do they say we are going to offer X ITA's this draw i.e. 3600 and then go down the pool until they get a cut off point then round it to the CRS score above so that's why a non round number is produced like 3534.

What I'm struggling from the figures to make out is why go in so high so quickly nearly 4000 ITA's is a huge number to offer and cannot be consistently offered.

Even if you splice in a few PNP draws which would make the ITA's probably around 1500 (guestimate) it still wouldn't allow an averaging out of more than 2200-2500 ITA's per month.

Unless there is a huge amount of people sitting in the 420-433 score then I just don't see where the CRS drop will stop. Purely guestimate based on the pool as of 3rd Jan 2016 - i'd imagine around 6000-8000 people with that score range currently.

If ITA's stay at around 3500 that would be cleared in the space of 3 draws.

Unless there is a huge amount of people in that range i.e. 10,000+ or a high % of new applicants are getting that score, then either 1 of two things will happen.

Either they will keep the CRS cut off around early 420's and therefore reduce the number of ITA's or continue to drop the points marginally by 1/2 each draw and then take a big intake of people once the 420 - 450 range fills up again.

I think that is the more they will likely make as appose to going towards 400.

Unfortunate for me but I think that makes more sense.

 

[picklee]

Carrying on stat conversation here

Unfortunately I cannot get any indication of how they decide a cut off point, do they say we are going to offer X ITA's this draw i.e. 3600 and then go down the pool until they get a cut off point then round it to the CRS score above so that's why a non round number is produced like 3534.

What I'm struggling from the figures to make out is why go in so high so quickly nearly 4000 ITA's is a huge number to offer and cannot be consistently offered.

Even if you splice in a few PNP draws which would make the ITA's probably around 1500 (guestimate) it still wouldn't allow an averaging out of more than 2200-2500 ITA's per month.

Unless there is a huge amount of people sitting in the 420-433 score then I just don't see where the CRS drop will stop. Purely guestimate based on the pool as of 3rd Jan 2016 - i'd imagine around 6000-8000 people with that score range currently.

If ITA's stay at around 3500 that would be cleared in the space of 3 draws.

Unless there is a huge amount of people in that range i.e. 10,000+ or a high % of new applicants are getting that score, then either 1 of two things will happen.

Either they will keep the CRS cut off around early 420's and therefore reduce the number of ITA's or continue to drop the points marginally by 1/2 each draw and then take a big intake of people once the 420 - 450 range fills up again.

I think that is the more they will likely make as appose to going towards 400.

Unfortunate for me but I think that makes more sense.

Here is the distribution of CRS scores for some ITAs:

(Full size: right-click, "View image")

Even if these numbers are incomplete, they give us a good representation of the probability of a certain score class. For example, nearly a third (32%) of all these profiles were invited with scores between 451-500 and only 0.4% were invited with less than 451!

However, the changes in November 2016 most likely affected the distribution of points. Unfortunately, IRCC has not yet released data for the recent period.

Here is the breakdown by economic program:

(Full size: right-click, "View image")

From these data, we can see that as CRS scores fall, IRCC is most likely drawing on FSW.

Data come from here: http://www.cic.gc.ca/opendata-donneesouvertes/data/IRCC_EEITA_0010_E.csv

 

[robb83]

Nice post ^ I like that

I'm copying what I put in the 420 - 425 thread


If the CIC target of 70,000 is to be believed and accurate (I presume they wouldn't want to go over)

With 30 draws this year at 3K a draw you'd be looking at 90,000 - I just don't see them doing that.

If the next drop is another 6 points, to me that means they have cleared out that range of CRS based on how many ITA's they decide to grant.

Taking the data there - currently majority above 434 has been cleared this year, which is approx. 20k ITA's so far - unfortunately according to that data as of Jan 2016 we can see there must have been a big increase in the amount of people in the pool.

If 3000+ ITA's are clearing 6 points at a time then you'd have to by all accounts predict it would keep dropping.

However they cannot keep giving up 3000+ ITA's and how much does the volume of people waiting at another 6 point drop equal, are there 3600 worth of ITA's up to 429 for example, and so on.

With the year end 2016 data we would be able to know based on current trends pretty much, whos CRS point would likely to get an invite and when.

 

[robb83]

Key piece of data missing is how the pool stood as of end of 2016.

If they released that then we could guestimate a decent trend.

With the 600 points going away they cannot keep points realistically over 450 otherwise it would be back to sub 2000 ITA's per draw easily even with the new points change.

The problem is there understanding just how many people are sitting above 400 points and how many new people are coming in week in week out into that 400 - 450 range.

 

[picklee]

The problem is there understanding just how many people are sitting above 400 points and how many new people are coming in week in week out into that 400 - 450 range.

There's no reason to believe that the number of new profiles coming in are any different from the distribution of scores above. For example, here is the same distribution expressed as a percentage of all ITAs issued:

(Full size: right-click, "View image")

 

[xpressentry]

Nice post ^ I like that

I'm copying what I put in the 420 - 425 thread


If the CIC target of 70,000 is to be believed and accurate (I presume they wouldn't want to go over)

With 30 draws this year at 3K a draw you'd be looking at 90,000 - I just don't see them doing that.

If the next drop is another 6 points, to me that means they have cleared out that range of CRS based on how many ITA's they decide to grant.

Taking the data there - currently majority above 434 has been cleared this year, which is approx. 20k ITA's so far - unfortunately according to that data as of Jan 2016 we can see there must have been a big increase in the amount of people in the pool.

If 3000+ ITA's are clearing 6 points at a time then you'd have to by all accounts predict it would keep dropping.

However they cannot keep giving up 3000+ ITA's and how much does the volume of people waiting at another 6 point drop equal, are there 3600 worth of ITA's up to 429 for example, and so on.

With the year end 2016 data we would be able to know based on current trends pretty much, whos CRS point would likely to get an invite and when.

The target is 70,000. Even if they have 30 draws and issue 3000+ invites, all 90,000 will not be successful in gaining PR. So they will always invite more than their target. Not to forget backlog from previous years and this years applicants going into the next year.

 

[robb83]

30% over their target to accommodate for people not managing to get PR not sure how accurate that will be however interesting point.

Even if we still take that as 3000 ITA's issued and depending on how many new people enter the pool with between 400 - 450 will affect how low CRS drops.

Shame they haven't told us precisely how they decide upon a cut off point.

 

[picklee]

30% over their target to accommodate for people not managing to get PR not sure how accurate that will be however interesting point.

The 2015 data show approximately 63% of applicants were approved. That translates to 0.91 ITAs per landed PR.

Even if we still take that as 3000 ITA's issued and depending on how many new people enter the pool with between 400 - 450 will affect how low CRS drops.

We can assume about 32% of new profiles entering the pool immediately after a draw will fall within the 451-500 range. So 101,928 eligible EE profiles were submitted between 01JAN2015 03JAN2016, which is 1,960 profiles per week. We could estimate that 627 of those would fall between 451-500.

The target is 70,000. Even if they have 30 draws and issue 3000+ invites, all 90,000 will not be successful in gaining PR. So they will always invite more than their target. Not to forget backlog from previous years and this years applicants going into the next year.

Also, each ITA is associated with 1.7 applicants, on average.

 

[ILoveBritishColumbia]

Nice statistical analysis, picklee!

With regard to the distribution of CRS scores, a more detailed breakdown can be extracted from the report titled “Refocusing Express Entry July 26 2016 Stakeholder Consultations” (I was unable to post the hyperlink, but the pdf will come up if you Google the title of this report. Incidentally, this report has lots of other interesting/useful info regarding past and future EE draws and criteria, as mentioned in other threads.)

On p.9 of this report, a bar graph with percentages of candidates in each ten-point range of core CRS scores is given (based on a March 28, 2016 snapshot of active candidates in the pool). The graph only gives a visual representation of percentages; I translated that into (estimated) actual percentages, and then also calculated an estimated number of candidates with that score based on a total of 57,500 active candidates in the pool as of May 2016 (p.4). That yields the following table for percentage and number of candidates with core CRS scores 370 and above:

Core CRS: % # of candidates with that score
370’s 8.1% 4658
380’s 6.8% 3910
390’s 5.4% 3105
400’s 5.7% 3278
410’s 5.3% 3048
420’s 3.9% 2243
430’s 5.2% 2990
440’s 4.6% 2645
450’s .8% 460
460’s+ .5% 288

The graph in this report shows the distribution of “core CRS scores”, that is, CRS scores once the extra points for a job offer or PNP have been removed. As such, it is useful for trying to factor out the influence of those extra job points (previously 600 pts, now 50-200 pts) since that has now changed.

Again, this table/graph and the stats it’s based on are not from 2016 year-end, nor do they reflect the recent CRS changes that give more points for Canadian education, etc. (like the stats in picklee's analysis). Undoubtedly the number of candidates with core CRS scores above 400 has now increased significantly, since a certain percentage of these candidates will now have received up to 30 more points. However, this snapshot does give the breakdown for relative number of candidates in each ten-point range at the time, which is potentially useful for statistical analyses/estimates/extrapolations/speculations/guesses.

 

[robb83]

The 2015 data show approximately 63% of applicants were approved. That translates to 0.91 ITAs per landed PR.

We can assume about 32% of new profiles entering the pool immediately after a draw will fall within the 451-500 range. So 101,928 eligible EE profiles were submitted between 01JAN2015 03JAN2016, which is 1,960 profiles per week. We could estimate that 627 of those would fall between 451-500.

Also, each ITA is associated with 1.7 applicants, on average.

So if we say amount of ITA's issued remains constant and for rounding purposes we call 700 profiles enter at 451-500 per week, we would say that inbetween each draw 700/1400 out of the 3500 are being used up by the 451-500 boundary.


Draw 51 Jan-04 2,902 468 700 2,202
Draw 52 Jan-11 3,334 459 700 2,634
Draw 53 Jan-25 3,508 453 1400 2,108
Draw 54 Feb-08 3,664 447 1400 2,264
Draw 55 Feb-22 3,611 441 1400 2,211
Draw 56 Mar-01 3,884 434 700 3,184


So from that last draw we would say that there were 3,184 in the pool with between 434 - 450 points

Interesting then that there would have only been 2,211 in the pool with 441 - 450

If that trend continues with a 6 point drop ITA's could be over 4,000

I just don't see the points dropping anymore otherwise we would be seeing almost 5000 ITA's a month.

*Another point ive considered whilst re-reading, would the pool be cleared out in terms of the points and above after the ITA has been issued?

So when there is a draw consecutive weeks, are people who were offered ITA from the previous draw still considered in anyway?

Anyone got any logic on the % of entries per week between 410 - 450?

 

[picklee]

With regard to the distribution of CRS scores, a more detailed breakdown can be extracted from the report titled “Refocusing Express Entry July 26 2016 Stakeholder Consultations” (I was unable to post the hyperlink, but the pdf will come up if you Google the title of this report. Incidentally, this report has lots of other interesting/useful info regarding past and future EE draws and criteria, as mentioned in other threads.)

Here's the link to that report for the peanut gallery:
https://c.ymcdn.com/sites/www.cerc.ca/resource/resmgr/government_relations/Express_entry_national_stake.pdf

On p.9 of this report, a bar graph with percentages of candidates in each ten-point range of core CRS scores is given (based on a March 28, 2016 snapshot of active candidates in the pool). The graph only gives a visual representation of percentages; I translated that into (estimated) actual percentages, and then also calculated an estimated number of candidates with that score based on a total of 57,500 active candidates in the pool as of May 2016 (p.4). That yields the following table for percentage and number of candidates with core CRS scores 370 and above:

Core CRS: % # of candidates with that score
370’s 8.1% 4658
380’s 6.8% 3910
390’s 5.4% 3105
400’s 5.7% 3278
410’s 5.3% 3048
420’s 3.9% 2243
430’s 5.2% 2990
440’s 4.6% 2645
450’s .8% 460
460’s+ .5% 288

The graph in this report shows the distribution of “core CRS scores”, that is, CRS scores once the extra points for a job offer or PNP have been removed. As such, it is useful for trying to factor out the influence of those extra job points (previously 600 pts, now 50-200 pts) since that has now changed.

+1 These percentages are helpful. I can't find any statistics on candidates with job offers, but we do know quite a bit about about PNP. In fact, if you look at the distribution of PNP in the colored chart, we see that the majority is between 900-1100. If we subtract 600, we can derive that the majority of PNPs have core CRS scores between 300-500, which aligns with the majority of candidates in your table above as well as the grey bar plot.

The fact that the grey bar plot is bi-modal suggests that we have two distinct populations: regular EE candidates and candidates with a +600 point bump from PNP/job.