--上面的数据特殊，改一下
declare @a table(riqi smalldatetime,         area  varchar(10),      info varchar(10))
insert @a select '2007-10-6',     '南京',          '房价' 
union all select '2007-10-6'     ,'上海',          '猪肉' 
union all select '2007-10-8'     ,'西安',          '工资' 
union all select '2007-10-8'     ,'南京',          '交通' 
union all select '2007-10-8'     ,'上海',          '美女' 
union all select '2007-10-9'     ,'上海',          '美女'

select area from @a a where riqi<'2007-10-31' group by area having(count(1)=(select distinct  count(distinct riqi) from @a where riqi<'2007-10-31'))
--result
/*
area       
---------- 
上海

（所影响的行数为 1 行）
*/