There is growing interest in so called "rehabilitation codes." This article will evaluate the risks in existing buildings, as opposed to new construction. There are two settings in which rehabilitation
codes are offered as a solution to two types of problems:

  • Vacant or abandoned buildings. These are seen as magnets for bad events, including but not limited to fires, resulting in a wider negative impact on their neighborhoods and even their communities.
  • "Under-utilized" buildings whose owners wish to improve their value and marketability through improvements and/or repurposing, such as changes in occupancy.

Vacant or Abandoned Buildings

There is a lot of statistical information on vacant housing and very little on vacancy rates in other kinds of properties.


Vacancy rates in housing have been rising since at least the early 1980s. Since the recent bursting of the housing bubble, the rate of increase in housing vacancies has accelerated. Vacancy rates around 9% in the early 1980s have given way to a new high of 14% in 2008. 1 Vacancy rates are generally higher in apartments than in other homes. In 2005, for example, all housing combined averaged a 13% vacancy rate. Single family homes (including manufactured homes) averaged a 9% vacancy rate, and the one- and two-family home category averaged a 9-12% vacancy rate. (The range reflects the fact that housing unit data combines buildings with 2-4 units. The upper rate applies if all 2-4 unit buildings are actually two-unit buildings, the lower rate if none are.) Meanwhile, apartments averaged a vacancy rate of 17%.


Rates of fires per thousand housing units are typically lower in vacant housing than in non-vacant housing. While the overall housing vacancy rate was 13%, the vacant-building percentage of all home fires was 5-6%. (This range reflects different assumptions about whether buildings under construction or demolition are counted in housing statistics and whether "idle" buildings are also vacant buildings.) The vacant-building percentages were 1% for civilian fire deaths and civilian fire injuries and a slightly higher 7-8% for direct property damage.

These lower percentages suggest that the risk of having a reported home fire in a vacant housing unit overall or weighted by severity of loss in monetary terms was roughly half the corresponding risk in a non-vacant housing unit.


The patterns are different in degree if one- and two-family homes are separated from apartments. While the overall vacancy rate in one- and two-family homes was 9-12%, the vacant-building percentage of one- and two-family home fires was 7-8% (and 8-9% for direct property damage), indicating a marginally lower fire risk than for non-vacant units. While the overall vacancy rate in apartments was 17%, the vacant-building percentage of apartment fires was 2% (and 3% for direct property damage), indicating a much lower fire risk than for nonvacant apartments.


A vacant housing unit will have less human activity, less powered equipment and less use of heat sources than a non-vacant housing unit. The only exceptions are cases of arson or playing with fire, where the absence of occupants means an absence of supervision. A vacant apartment in an otherwise occupied building may have the benefits of vacancy less use of energy in the unit and few of the problems.


Fire statistics distinguish between vacant secured buildings and vacant unsecured buildings. Abandoned buildings probably constitute part of the second group. There are roughly as many fires in vacant secured housing as in vacant unsecured housing. However, there is no data to split vacant housing units between secured and unsecured. Therefore, there is no way to check the hypothesis that secured vacant housing units have less fire risk than unsecured vacant housing units.

The relative size of the intentional fire problem in the two types of vacant housing units can be distinguished from all housing units combined. Intentional fires accounted for 8% of all home fires in 2003-2006, 35% of vacant secured home fires and 61% of vacant unsecured home fires. Intentional fires accounted for 8% of all one- and two-family home fires in 2003-2006, 36% of all vacant secured unit fires and 61% of all vacant unsecured unit fires. Intentional fires accounted for 6% of all apartment fires in 2003-2006, 24% of all vacant secured apartment fires and 62% of all vacant unsecured apartment fires. In every type of housing, vacancy increases the relative risk of having an intentional fire versus the risk of having another type of fire, and the absence of security further increases that relative risk of fire occurring.


Putting this all together, vacancy does not make a housing unit more likely to have a fire, but it may make it more likely to have an intentional fire, and abandonment probably raises both the risk of intentional fire and the overall risk of fire, at least for one- and two-family homes. On the other hand, abandonment means there is no owner seeking to rehabilitate the property.


If there is a vacant housing unit with an identifiable owner, local authorities can compel the owner to secure the property until it is sold, and that may take care of their interest in the situation. If there is a vacant housing unit without an identifiable owner an abandoned housing unit then local authorities will want to secure it and/or divert it to some acceptable use. That could mean transferring ownership to someone who will take on the costs of converting the property to a useful form. It could mean demolishing the property.


Thanks to the housing bubble, there is more housing than households seeking housing. Homelessness and foreclosures reflect increased difficulty in matching housing with households, but that is not the same as lack of supply. Vacancy is also probably on the rise for other types of buildings, such as mercantile properties, offices and hotel rooms, but again, there are probably enough units for those who can afford them and for those who would rent or buy if they could afford them.


In a situation like this, the community may not have an interest in keeping every building occupied. To reduce "sprawl," there may be a desire in densely populated cities for more open spaces than buildings. With an appropriate campaign of demolition and landscaping, one may add more value by removing buildings than by rehabilitation.


"Under-Utilized" Buildings

Suppose there is an old building that is not compliant with the current building code. Suppose it is kept in good repair, but it lacks the size or the features found in competing properties in today's market. Or suppose it is no longer viable for its original purpose (e.g., a factory) but might be viable for a different purpose (e.g., a mall for craft stores). An owner could be looking at a substantial cost to make the changes needed to make the property either competitive in its old market or viable in its intended new market. If there are enough changes, provisions may be triggered that require the building to meet new construction requirements. The business case may not look so workable.


A building code is the expression of a community's shared views on what it takes to make a building acceptably safe for a particular use safe for neighbors and safe for occupants, most of whom are not owners, including employees, guests, patients, customers, students and even inmates. (The reason there are property maintenance codes, like NFPA1, Fire Code,2 is that the same community realizes that the level of safety provided during construction may not remain in place indefinitely without maintenance.)


The cost of providing a defined level of safety can be much higher for an existing building than for a new building, and the gap will be different for some features than for others. Changes to locks may be easy and inexpensive. Changes to detection and alarm systems may be relatively inexpensive. Changes to sprinklers may be more expensive. Changes to basic construction would be the most expensive and least practical of all.


The reason there are different provisions for existing buildings as in NFPA 101, Life Safety Code,3 and NFPA 5000,4 Building Construction and Safety Code, Chapter 15 is that the same community that set levels of acceptable safety also has decided that the margin of additional safety is not large enough to justify the much greater additional cost involved in retrofitting every element required for new construction. Rehabilitation codes are an expansion of that concession. Instead of having two codes one for existing buildings and one for new construction or any change of sufficient size to existing buildings rehabilitation codes create multiple sets of requirements, depending on the type and magnitude of changes. The U.S. Department of Housing and Urban Development (HUD) called this "adaptive reuse" and supported it in 1995 through what they termed Nationally Applicable Recommended Rehabilitation Provisions (NARRP).5


There are a couple other precedents for this kind of thinking. One is codes for historic buildings and cultural resources. These properties have non-cost reasons for avoiding new-construction requirements. The other is performance-based design. This is a systematic procedure for demonstrating equivalent performance. Both precedents have benefits in the area of rehabilitation.


Rehabilitation codes at their best are well-considered variations on the rules used to assure that public values are properly included in private decisions that affect everyone. They often are more complex, but for good reason. Many of the arguments used to promote rehabilitation codes do not withstand close examination. Vacant buildings are everywhere and do not attract fires. Abandoned buildings do, but those aren't the buildings whose owners are seeking to rehabilitate. Under-utilized buildings are a problem for their owners but not necessarily for the larger community. As with the professional sports team owners who try to sell the community on paying for a big new facility, the economic case for relaxed rules for building conversions or enhancements may not withstand close examination.


All of these caveats are reasons why the published rehabilitation codes deserve informed support. The volunteers who wrote these codes took seriously the upside potential and the downside risk of introducing this approach, and they have installed both guidance to make it possible and circuit-breakers to keep it safe.


John Hall is with the National Fire Protection Association.



  1. Ahrens, M., "Vacant Building Fires," NFPA Fire Analysis & Research Division, April 2009.
  2. NFPA 1, Fire Code, National Fire Protection Association, Quincy, MA, 2009.
  3. NFPA 101, Life Safety Code, National Fire Protection Association, Quincy, MA, 2009.
  4. NFPA 5000, Building Construction and Safety Code, National Fire Protection Association, Quincy, MA, 2009.
  5. Nationally Applicable Recommended Rehabilitation Provisions, U.S. Department of Housing and Urban Development, Washington, DC, 1997.