NOT IN Returns Unexpected Results

Question

I am running the following queries in MS SQL Server (version 2016 if that matters).

SELECT DISTINCT Job_Number
FROM table_A
WHERE Job_Number IS NOT NULL

returns 2376 rows

SELECT *
FROM table_A
WHERE Job_Number IN (SELECT DISTINCT [House Bill] FROM table_B)

returns 137 rows

However,

SELECT *
FROM table_A
WHERE Job_Number NOT IN (SELECT DISTINCT [House Bill] FROM table_B)`

returns 0 rows. I would expect this would return 2239 rows (i.e. 2376 - 137).

Is there any fundamental concept of NOT IN that I have yet to understand here?

score 2 · Accepted Answer · answered Oct 12 '18 at 11:57

NOT IN will fail to return results if there are any NULL values returned by your following query: (SELECT DISTINCT [House Bill] FROM table_B)

My suggestion is to modify your query to:

(SELECT DISTINCT [House Bill] FROM table_B WHERE [House Bill] IS NOT NULL)

The reason this happens is because whenever a comparison is made between a row in table_A with NULL value in table_B the logical result is UNKNOWN.

There's more information about this:

1 - here and

2 - here

score 0 · Answer 2 · answered Oct 12 '18 at 11:52

0

The IN() clause ignores NULL values.

answered Oct 12 '18 at 11:52

juergen d

195,137
36
275
343

score 0 · Answer 3 · answered Oct 12 '18 at 11:54

The possible reason would be null values in tableb in [House Bill] column :

The correct way would be to use NOT EXISTS OR EXISTS instead of NOT IN AND IN clause :

SELECT a.*
FROM table_A a
WHERE NOT EXISTS (SELECT 1 FROM table_B b WHERE a.Job_Number = b.[House Bill]);

score 0 · Answer 4 · answered Oct 12 '18 at 11:57

NOT IN returns no rows at all if the subquery returns NULL -- even once.

For that reason, I strongly recommend using NOT EXISTS. This behaves as you expect:

SELECT a.*
FROM table_A a
WHERE NOT EXISTS (SELECT 1 NFROM table_B b WHERE a.Job_Number = b.[House Bill])

NOT IN Returns Unexpected Results

4 Answers4