Table 4.  Estimates of the Total Housing Inventory for the
                United States:  Third Quarter 2004 and 2005

(Estimates are in thousands and may not add to total, due to rounding)


                                    Third     Third          90-Percent
                                   quarter   quarter   Confidence Interval      2005
Type                                 2004      2005    of 2005       of       Percent
                                   estimate  estimate  estimate  difference  of total

All housing units.................  122,373   124,119       (X)        (X)       100

  Occupied........................  106,870   108,431       330        301        87
    Owner occupied................   73,772    74,588       631        432        60
    Renter occupied...............   33,098    33,843       536        413        27

  Vacant..........................   15,503    15,688       351        311        13
    Year-round vacant.............   11,848    11,854       343        297        10
      For rent....................    3,798     3,773       172        193         3
      For sale only...............    1,321     1,481        92        113         1
      Rented or sold,
      awaiting occupancy..........    1,050     1,191        75        103         1
      Held off market.............    5,679     5,409       239        210         4
        For occasional use........    1,955     1,799       140        124         1
        Temporarily occupied
        by persons with
        usual residence elsewhere.    1,026     1,020       106         92         1
        For other reasons.........    2,698     2,590       167        147         2

    Seasonal vacant...............    3,655     3,834       225        195         3
    aA 90-percent confidence interval is a measure of an estimate's reliability.  The larger the 
confidence interval is, in relation to the size of the estimate, the less reliable the estimate.
         (X)  Not Applicable.  Since the number of housing units is set equal to an independent national 
measure, there is no sampling error, and hence no confidence interval.
NOTE:  Since first quarter 2003, the Current Population Survey/Housing Vacancy Survey (CPS/HVS) estimates 
have been controlled to independent housing unit counts.  Doing so should make the CPS/HVS estimate of housing 
units more comparable to other Census Bureau housing surveys.