-- explain3.txt

+----+-------------+----------+--------+---------------+---------+---------+------------------------+------+-------------+
| id | select_type | table    | type   | possible_keys | key     | key_len | ref                    | rows | Extra       |
+----+-------------+----------+--------+---------------+---------+---------+------------------------+------+-------------+
|  1 | SIMPLE      | employee | ALL    | PRIMARY       | NULL    | NULL    | NULL                   |  118 |             |
|  1 | SIMPLE      | person   | eq_ref | PRIMARY       | PRIMARY | 5       | classfiles.employee.id |    1 | Using where |
+----+-------------+----------+--------+---------------+---------+---------+------------------------+------+-------------+

[Results may vary in the details, but should be similar]

MySQL will scan through all 118 "employee" records.  For each of these records, it will then use the PRIMARY key on
the "person" table to look up the associated row.  There will thus be 118 lookups, and MySQL expects each lookup to
return a single row.  The results from each lookup will then be examined to see if they pass the WHERE clause.

MySQL realizes that without an index on the "area_code" column, it would have to scan through more than 1000 rows in
the "person" table to find all the NULL area codes, and this would be slower than scanning through only 118 rows in
the "employee" table.