Rewrite T-SQL from using IN to using JOIN/WHERE

Question

How can I rewrite this t-sql to avoid using IN - and use joins in stead?

--Members with membership in 2008 but not in 2009, using IN and NOT IN 
SELECT * FROM @Members 
WHERE MemberId NOT IN 
 (SELECT MemberId FROM @Memberships WHERE [Year] = 2009) AND 
 MemberId IN (SELECT MemberId FROM @Memberships WHERE [Year] = 2008 )

I want to find people with a membership in 2008, who do not have a membership in 2009. The correct solution in the below example is George and Bill

Something like this...:

--Members with membership in 2008 but not in 2009, using IN and NOT IN 
SELECT * FROM @Members m
 INNER JOIN @Memberships msOld ON msOld.MemberId = m.MemberId AND [Year] = 2008
 PLAYTHEMOFFKEYBOARDCAT JOIN @... HUh?

Complete schema to play with. :)

DECLARE @Members TABLE

(
  MemberId INT,
  Name VARCHAR(10)
)

INSERT INTO @Members VALUES (1, 'Joe')
INSERT INTO @Members VALUES (2, 'Mike')
INSERT INTO @Members VALUES (3, 'George')
INSERT INTO @Members VALUES (4, 'Bill')
INSERT INTO @Members VALUES (5, 'Kyle')
INSERT INTO @Members VALUES (6, 'Greg')

DECLARE @Memberships TABLE
(
  MembershipId INT,
  MemberId INT,
  [Year] INT
)

INSERT INTO @Memberships VALUES (1, 1, 2009)
INSERT INTO @Memberships VALUES (2, 1, 2007)
INSERT INTO @Memberships VALUES (3, 1, 2006)
INSERT INTO @Memberships VALUES (4, 2, 2008)
INSERT INTO @Memberships VALUES (5, 2, 2009)
INSERT INTO @Memberships VALUES (6, 3, 2007)
INSERT INTO @Memberships VALUES (7, 3, 2008)
INSERT INTO @Memberships VALUES (8, 4, 2008)
INSERT INTO @Memberships VALUES (9, 5, 2007)
INSERT INTO @Memberships VALUES (10, 5, 2006)
INSERT INTO @Memberships VALUES (11, 5, 2005)
INSERT INTO @Memberships VALUES (12, 6, 2007)

--SELECT * FROM @Members m INNER JOIN @Memberships ms ON m.MemberId = ms.MemberId ORDER BY [Year]

--Membership in 2008
SELECT MemberId FROM @Memberships WHERE [Year] = 2008

--Members with no membership in 2009
SELECT * FROM @Members WHERE MemberId NOT IN 
    (SELECT MemberId FROM @Memberships WHERE [Year] = 2009)

--Members with membership in 2008 but not in 2009, using IN and NOT IN 
SELECT * FROM @Members WHERE MemberId NOT IN 
    (SELECT MemberId FROM @Memberships WHERE [Year] = 2009) AND MemberId IN (SELECT MemberId FROM @Memberships WHERE [Year] = 2008 )

--Members with membership in 2008 but not in 2009, using IN and NOT IN 
SELECT * FROM @Members m
    INNER JOIN @Memberships msOld ON msOld.MemberId = m.MemberId AND [Year] = 2008

Answer 1

--Members with membership in 2008 but not in 2009, using JOINs
SELECT M1.*
FROM @Members M1
JOIN @Memberships MS08 ON (M1.MemberId = MS08.MemberId AND MS08.Year = 2008)
LEFT JOIN @Memberships MS09 ON (M1.MemberId = MS09.MemberId AND MS09.Year = 2009)
WHERE MS09.MemberId IS NULL

Answer 2

It's better as an EXISTS.

EXISTS will never be worse then an OUTER JOIN, but is often better. To my eyes, it's also a "cleaner" query and is more easily understood.

SELECT
    M1.*
FROM
    @Members M1
WHERE
    EXISTS (SELECT *
             FROM @Memberships MS08
             WHERE MS08.Year = 2008 AND M1.MemberId = MS08.MemberId)
    AND
    NOT EXISTS (SELECT * 
             FROM @Memberships MS09 
             WHERE MS09.Year = 2009 AND M1.MemberId = MS09.MemberId)

Answer 3

SELECT m.MemberId FROM @Members m JOIN @Memberships ms ON (m.MemberId = ms.MemberId)
    WHERE ms.[Year] BETWEEN 2008 AND 2009
GROUP BY m.MemberId
HAVING MAX(ms.[Year]) = 2008