Manufactured Housing Is a Good Source of Unsubsidized Affordable Housing -  Except When It’s Not: Q&A on Eight Key Policy Topics (Part 2)

Research & Policy | April 3rd 2023 | Donald H. Layton

Manufactured housing community

Manufactured housing (MH) has recently taken on a high profile among affordable housing advocates, including in the Biden administration. MH’s supporters consider it naturally low-cost due to the efficiency of factory production, rather than via government subsidy.  As a result, the search for ideas to increase the usage of MH has been building given the critical affordable housing shortage that has grown for many reasons over at least the last decade.

As described in Part 1 of this three-part series, MH – a term which in this paper refers to the factory-produced structure that is later sited on a residential lot – accounts for approximately six percent of owner- and renter-occupied housing in the U.S., with a much greater market share of 14 percent in rural geographies.  MH also has more complex economics than conventional owned or rented housing, particularly with respect to the significant market share of MH where the structure is owned by its resident but the underlying land is rented from someone else.  The challenge in such cases is that the MH resident does not get to enjoy the two significant social benefits of homeownership that have long been used to justify government intervention to subsidize and support it:  building net worth to support retirement and the next generation, and avoiding landlord-generated housing instability. 

Nevertheless, housing advocates describe MH as a relatively simple market that can provide much more affordable housing units if only two straightforward things were done.  First, have the four government mortgage agencies1 strongly support2 “chattel” loans (i.e., loans secured only by the MH structure, excluding the underlying land); and second, look to the states to override what is characterized as unjustified and unfair zoning and other land use rules that restrict MH structures generally and MH communities (MHCs) more specifically.  Additionally, some advocates also call for preventing institutional investors – and maybe even all for-profit investors – from being MHC owners to avoid excessive housing-related family instability, up to and including the extreme case of forced removal of the MH structure from the underlying land.

Part 2 examines eight key policy topics, via a Q&A format, that must be accurately understood to develop effective policy that can expand affordable housing via MH.  These topics reveal that MH is indeed a very complex type of housing.  Thus, as will be addressed in Part 3 of this series, recommendations to increase it as a source of affordable housing are carefully targeted and sometimes just as complex. 

The Eight Key MH Policy Topics

Q1:  How does the durability and quality of MH structures compare to site-built ones?

A:  MH structures and site-built homes are both produced in a wide range of durability and quality commensurate with their range of price points.3  The average MH structure has materially less durability and quality than the average site-built home, while the best MH seems to come close to matching the site-built average. 

The MH industry historically has produced what it calls three levels of quality,4 where the highest has finishes roughly equal to those found in the average site-built home, while the lowest is very much oriented towards affordability by using lower-quality materials and construction techniques.5 (More recently, the industry has introduced the “CrossMod” type home, a hybrid where an MH structure has additional site-built features – such as an enclosed garage – to get beyond the narrow design constraints that apply to a truck-pulled MH unit.  This will increase MH manufacturers’ competitiveness at the higher end of its price range, although it is too soon to determine if significant volume will result.6)  Nevertheless, all newly built MH units, even those with the very low cost, meet the building standards established by HUD, including energy efficiency (which leads to lower monthly operating costs).  However, the concept of “quality” is somewhat vague – often meaning mainly on-the-surface finishes (e.g., stone vs. Formica countertops).  In my view, the concept of quality that is focused instead on durability (or expected useful lifespan) is more important, i.e., how long an MH structure – due to the quality of its engineering and the longevity of structural materials and major systems, such as heating – is expected to last.

In terms of durability, the lifespan of an average and typical newly constructed MH unit is cited by HUD as being from 30 to 55 years.7  The typical site-built house can easily last more than 100 years, which is a much longer span over which to amortize the purchase price of a residential structure.  Thus, apples to apples, the average MH structure’s durability is considerably less than that of an average site-built home.  However, as discussed in Part 1, the durability of any residential structure will be inextricably bound up in the ability and willingness of the homeowner to spend money on routine maintenance and repair, as well as making major capital investments as needed from time to time.  Of relevance to this is that the median annual household income of MH structure owners is about $35,000, just half of the median annual income of all site-built homeowners, and over one-quarter earn less than $20,000 annually, where about a third of site-built homeowners earn less than $50,000 annually.8 This means that, on top of any difference in new-build quality and durability,9 MH will likely have a measurably shorter lifespan than site-built homes simply because MH owners have materially less income to support maintenance and any needed significant investments.

Thus, the average MH is not as durable and high quality as the average site-built home.  But a recent study by the Urban Institute10 compared the appreciation in prices over a long period of time of a pool of what it calls “high-end” MH11 (which I view as another term for the “best” MH) versus the site-built average.  The result showed that the best MH came close to appreciating the same.  Using that price appreciation as a reasonable proxy for a measure of quality and durability, one can therefore conclude that the best MH has quality/durability approximating that of the site-built average.

Q2:  How much less expensive is MH when compared to site-built homes, and what portion of that is due to the efficiency of factory production?

A:  The average MH structure is less expensive than the average site-built structure, probably in the range of 40 percent less per square foot.  This is due partially to the efficiency of MH factory production and partially to the average MH structure’s lower durability and quality versus the average for a site-built structure.  It is unknown how much each factor contributes to the lower per-square-foot cost. 

The most frequently cited reference I have seen to the cost per square foot for MH comes from the Manufactured Housing Institute (MHI), an industry organization, which reports it to be 50 percent lower than a traditional home.12  This calculation requires estimates,13 and as MHI is an industry organization, it is reasonable to assume its methods probably produce optimistically high results.14  A separate calculation by the Urban Institute, an MH supporter, from 2018 claims manufactured homes are 35 to 47 percent cheaper per square foot than site-built homes.15 Thus, I would estimate that MH is actually around 40 percent less expensive.

The comparisons are done on a per-square-foot basis, a reasonable method to produce a more apples-to-apples result.  However, while adjusting for square footage is necessary, as site-built homes are on average larger than manufactured ones, it is not sufficient to properly compare costs.  The reality is that the average site-built home also has, as discussed above, longer durability and also better-quality finishes and features than the average MH structure.

It is unknown how much of my estimated 40 percent lower per-square-foot cost is due to the efficiency of factory production rather than just reflecting somewhat lower quality and durability.  Lacking any real evidence, my personal guesstimate is that factory production reduces the cost by 20 to 25 percent, with the rest of that reduction due to lower average quality and durability.  I note that this 20 to 25 percent estimate is still a considerable amount, even if less than the MHI-claimed 50 percent.

Q3:  How much does an MH structure appreciate or depreciate over time, especially in comparison to site-built homes? 

A: When the term “manufactured housing” refers specifically to owning an MH structure but renting the underlying land (which is the most common configuration), over the long term the owned MH structure depreciates. This is after adjusting for inflation and before the benefit of expenditures made to maintain it.  When the same term refers to owning an MH structure plus also the underlying land (especially when they are legally and physically joined together, which happens about one-fifth of the time), the combination of the two appreciates, although on average not quite as much as do site-built homes.   

As already described in Part 1, all residential structures – whether MH, site-built, modular, or even multifamily –are man-made and thus depreciate over time. MH generally does so a bit faster than other types due to its shorter lifespan.  However, this can be hard to discern at times due to inflation and property values benefitting from continuing expenditures to maintain them, and especially during strongly price-appreciating markets (such as the pandemic-distorted years of 2020 and 2021); nevertheless, it is true over the long-term.  It is instead the land upon which a residential structure – whether MH or any other type – sits that provides the increase in real value for the typical homeowner.16 

Thus, site-built housing, and virtually any other type of housing, including MH, is considered a good long-term investment – broadly believed to be the main way in which the typical family builds up net worth. However, this is true only when it consists of structure-plus-land together.  What makes homeownership for MH unique is how often structure and land ownership are separated; for all other housing types, this only happens rarely. In short, when it comes to long-term appreciation, it matters whether the “home” consists of structure-plus-land or just a structure; the type of structure has little to do with it.

Nevertheless, the advocacy I have found with respect to MH demonstrates some confusion regarding this fundamental.  Here is one example: 

“You may have heard that, over time, manufactured homes lose value compared with traditional homes.  The good news is that modern manufactured homes seem to have overcome this.  According to a study by the North Carolina Manufactured and Modular Homebuilders Association, if a manufactured home is classified as real property and affixed to a permanent foundation, it tends to appreciate in value at the same rate as nearby traditional homes.”17 [Emphasis added]

I have frequently found this confusing approach on the topic of appreciation versus depreciation.  It starts with referring to “manufactured homes” in an ambiguous way, being unclear if the underlying land is owned or rented. It then uses as evidence data that comes only from the one-fifth of cases where the underlying land is owned and joined legally and physically.  It’s thus easy to see how many readers could end up thinking appreciation occurs even when the underlying land is rented rather than only when it is owned.

In fact, even an MH supporter organization as well-known as the Urban Institute has engaged in this confusing type of argumentation.  As an example,  an article they published titled “New Evidence Shows Manufactured Homes Appreciate as Well as Site-Built Homes,”18 talks in the same ambiguous manner about MH (i.e., without any qualification as to whether the underlying land is owned or rented), and how the evidence shows that, based on data from 1995 to 2018,  it appreciated at 3.4 percent per year, almost as much as site-built homes that appreciated at 3.8 percent per year.  However, not until two-thirds of the way through the article is it revealed that the data for the calculations is restricted only to MH that includes underlying land ownership that is legally and physically joined to the manufactured structure.  Again, the average reader, who might not get two-thirds of the way through the article or even possibly much past the headline, could easily end up believing that MH structure-only homes appreciate when that is simply not the case. 

Another source of misleading conclusions, whether intended or not, is based upon selecting the recent pandemic-distorted period of time to compare prices, as seen in a 2021 Washington Post article,19 where an example from 2019 to 2021, possibly the most rapidly increasing home price period in modern history,20 is given.

I found additional advocacy approaches designed to support a claim that MH structures will appreciate much like site-built structure-plus-land housing.  In each case I examined, I found to my satisfaction that the claim did not hold up.  My conclusion that an MH structure depreciates drives a different set of policy conclusions than proposed by those who believe it appreciates, as will be described in Part 3.

At this point, we must expand the analytical framework developed in Part 1.  That framework so far is summarized as follows:

  • residential land is, on average, a generally long-term (GLT) appreciating asset,
  • a residential structure (whether MH or any other type) is a GLT depreciating asset,
  • the average typical land-plus-structure home has a track record of net appreciation in value as well (i.e., the land’s appreciation is greater than the structure’s depreciation) where the asset values are adjusted for inflation and also before considering expenditures that maintain a home’s value (i.e., repairs, maintenance, and major capital improvements). 

The complex and unusual economics of MH stems from there being main three ownership configurations,21 whereas site-built housing almost always involves straightforward renting or owning.  This more nuanced framework will help us end the ambiguity and confusion that comes from using the unmodified term “manufactured housing.” The three configurations are:

  • MH rental (“rent-rent MH”), where both the structure and the land are owned by one or two landlords and rented to the structure’s resident.22  The economics are the same as for any other single-family or multifamily rental situation.  Except, of course, that the lower cost due to manufacturing efficiency – earlier estimated to be 20 to 25 percent – can hopefully be passed on through landlords in the form of lower rents to support those with the lowest incomes.  It is a shame more policy attention has not been paid to this configuration, which accounts for five percent of all rental properties. This vacuum will be discussed further in Part 3.
  • MH structure-and-land ownership (“own-own MH”), where both the structure and land are owned by the resident.  Sometimes the two are joined together legally and physically and with a proper foundation so that it can also be considered, together, “real property.”  This real property own-own MH qualifies for a traditional mortgage; all other own-own MH does not.  Own-own MH has economics similar to those of a site-built home, with just a bit faster average depreciation of the structure.
  • MH structure ownership with rental of the underlying land (“own-rent MH”).23  This configuration, neither full rental nor full ownership, is a hybrid between the two.  It is at the heart of the policy debate about MH, especially because it seems to be the most common type of MH residency.  To repeat the conclusion from Part 1:  Because the GLT depreciating structure is owned while the GLT appreciating land is rented, it does not produce, on average over the long term, any buildup of real net worth for the typical resident family to support retirement or to help the next generation.  Additionally, the resident household is still subject to landlord-generated housing instability, including the extreme case of forced removal.  Thus, in my view, it is inadvisable to refer to own-rent MH broadly as being homeownership given how that term is commonly understood – it is at best marginally so.24

The identification of the own-rent MH case as only marginally being true homeownership is a major conclusion of this series of papers and is perhaps the most important difference between my views and those of MH supporters.  It has significant policy implications, as will be discussed in Part 3.

Q4:  MH structure-only (i.e., chattel) loans have much higher interest rates and worse terms than traditional mortgages.  If the two GSEs, or other government mortgage agencies, began to provide such loans, would it produce much more borrower-friendly rates and terms?

A:  The GSEs or other government mortgage agencies decisively entering the structure-only MH loan market would produce, at most, only a modest benefit to borrowers.  As a reminder, the roughly one-fifth of MH residents who have real property-MH ownership, as defined above, already qualify for a conventional 30-year mortgage, just like an owner of a site-built house.

Structure-only loans currently carry much worse interest rates and somewhat worse terms than traditional mortgages.  One recent estimate25 is that interest rates are very roughly about 4 percentage points higher, and the final maturity is typically closer to 20 years, not 30.  There are in fact three proper reasons why this occurs: 

  • One, the credit risk for structure-only loans is higher since, by definition, only the GLT depreciating structure can serve as collateral, whereas for a traditional mortgage the collateral consists of both the structure and the GLT appreciating underlying land.  This will also specifically impact the final maturity, making it shorter than 30 years, as the structure-only collateral does not provide enough protection in the out-years, while the underlying land does.26 
  • Two, as the typical structure-only borrower has significantly lower income than the typical mortgage loan borrower (as previously described), the structure-only loan has significantly increased credit risk.   
  • And three, perhaps the most important reason is that the structure-only loan does not get to participate in, and therefore benefit from, the particularly low rates and friendly terms27 of the single-family agency mortgage-backed securities (MBS) market, known colloquially as the “TBA” (to-be-announced) market.28 This market is unique to first-lien residential mortgages. 

This third reason requires further, and somewhat detailed, explanation. The borrower of a conventional mortgage benefits tremendously from the evolution of the TBA market, which was not centrally planned or created by legislation but evolved over several decades as a collaboration between government mortgage agencies and the bond trading and investing community.  The TBA market is highly unusual for a consumer loan because of its noticeably low rates and very borrower-friendly terms based on (1) government-supported guarantees to investors against credit loss on the underlying loans (for which a fee is charged), and (2) extraordinarily high liquidity, near that enjoyed by the U.S. Treasury markets, reflecting both the TBA market’s extraordinarily large size and its unusually high standardization.29  It is not that the MH structure-only loan market is being discriminated against in having worse terms than the TBA market; it is rather that the extreme size (about $8 trillion is currently so securitized) and unusually high standardization of the conventional mortgage business, along with the involvement of the four government mortgage agencies, have delivered a windfall of lower rates and unusually borrower-friendly terms to the typical home first-lien mortgage borrower.  I note no such market exists for auto loans, credit card loans, second mortgage loans, or business loans – it is unique to traditional first-lien mortgages.30

The question then becomes whether this exceptional TBA market can also include structure-only MH loans.  The answer is a clear “no” – they are simply too different from conventional mortgages (e.g., having only a structure as collateral) to be considered in the highly standardized TBA market.  And no dedicated and specialized TBA market could emerge just for structure-only loans – that market would be far too small and does not have the necessary extremely high standardization found in the TBA market.  Also, note that the government does not control what does and does not qualify for the TBA market – that is controlled by mortgage securities market participants, as they are the ones who take the related risks.31  In their own self-interest, those participants are extremely cautious in ensuring that the market’s ultra-standardization and near-U.S. Treasury trading liquidity are not compromised.  

So, perhaps the government mortgage agencies getting involved in structure-only MH loans could take some other form, such as doing custom securitizations,32 potentially delivering modestly lower rates and terms due to the agencies’ lower cost of capital and imprimatur?   I agree that such a modest benefit may be possible, but in my view, the rates would still be significantly higher than for a standard mortgage given both the higher credit risk and not having the major benefit of the TBA market. 

On top of this, even if a modest benefit were obtained, it is unclear how much of this benefit would actually make its way to borrowers.  This is because, being secondary firms by law, the government mortgage agencies are only allowed to deal directly with primary market lenders –in this case, the relatively small number of MH structure-only lenders.  I note that the proposals advocating for the government agencies to enter the structure-only market do not include a promise (or enforceable mechanism) by primary market lenders to pass any resulting benefit through to actual MH residents, as opposed to significantly increasing their own profit margins.  Thus, it is totally unknown what portion of this modest benefit would reach its intended beneficiaries.33

Given the discussion herein of whether the own-rent MH market deserves policy support comparable to that given to true homeownership situations (due to the problematic side-effect that it does not deliver the two main benefits of homeownership), it is therefore unclear what good policy should be in this case.  This will be further addressed in Part 3. 

Q5:  Does the 14 percent market share of MH in rural areas being so much higher than the national 6 percent average mean that MH’s potential to provide additional affordable housing is also concentrated in rural and other low-density geographies?

A:  This alignment of MH with low-density geographies – such as rural areas, exurbs, and small cities – is significantly due to housing economics. (It is also possible, although not certain, that zoning restrictions play a material role, as further discussed below.) This inevitably leads to the conclusion that developing policy for MH in such geographies makes sense.

There are two sound economic reasons why MH policy initiatives should be aligned with low-density geographies.  First, site-built construction in less dense geographies is burdened with extra costs.   The typical site-built home usually requires a general contractor to be able to access a large web of skilled sub-contractors (plumbing, electrical, HVAC, windows, rough carpentry, fine carpentry, foundations, roofing, etc.), each of which in turn requires a large population to support an adequate volume of business.  That level of population density is found in urban and close-in suburban areas, but in rural or other less-dense areas there will be fewer such sub-contractors – producing less competitive pricing – and also the extra cost associated with much greater distances for them to travel. This will, of course, directly raise the cost of site-built construction. Thus, factory-built housing (whether MH or modular) additionally gains a competitive advantage in less-dense geographies.34  Similarly, much housing, especially with lower price points, is produced today via large-scale developments, with hundreds if not thousands of homes.  The economy of scale this generates is considerable but not available in low-density geographies, which cannot support such development volumes.

Second, the low cost of MH structures aligns with the low cost of land in less-dense geographies and is inconsistent with the high cost of land in urban cores and close-in suburbs.  Housing developers do not invest in purchasing high-cost land and then put on it low-cost structures (whether MH or any other type35) as the potential returns on their investment would just be too constrained.  Similarly, working-class families are unlikely to use their limited funds to purchase high-cost land and then put on it a low-cost MH structure to live in.36 For the large dollars invested, this just produces too little livable space for the family’s needs.   

The exception to the alignment of MH with less-dense geographies may be MHCs, where the high cost of land in denser geographies is offset by having only a small amount of underlying land per MH unit.   

The policy implications of this geographic alignment, and the possible MHC exception, will be discussed in Part 3.

Additionally, where MH is often compared to site-built housing in terms of cost, there are other methods of developing affordable housing.  Modular housing, also factory produced, is considered more expensive than MH37 but still less expensive than site-built homes, for example.  Apartment houses in rural areas are relatively infrequent, but they are a low-cost alternative to single-family homes, either on a rental or home ownership basis.  This will be further explored in Part 3 as well.

Q6: Are zoning and other land-use (ZLU) restrictions, which exist in many local communities to significantly limit MH (including both single-family MH38 and MHCs), creating an unreasonable and unwarranted major barrier to MH providing more affordable housing units?

A:  It is not really known how much ZLU restrictions against MH impede the creation of affordable housing units.  Additionally, where there are such restrictions, they are not always unreasonable. 

Based on my research, it is clear that ZLU restrictions against MH are found in many communities.  What is not clear, however, is how much they are mainly or mostly found in the urban or dense suburban geographies where single-family MH would not be built in the first place due to the economics involved – which means that those restrictions may not be particularly relevant. By contrast, the rural market share of MH at 14 percent suggests that such barriers do not widely exist in less dense geographies.  This is a topic for further research to get beyond generalities and into specifics.

It is also clear that, in the judgment of the marketplace, rather than being totally unwarranted because of the advocacy claim that MH is “just as good as site-built housing,” some barriers to MH in more dense geographies make sense to local government officials due to their potential to lower property values over time.39  This overlaps with the larger policy discussion about the trade-off between typical locally-driven zoning in more dense geographies versus the need for more housing, a topic outside the scope of this paper.  

Again, MHCs may be an exception, with barriers in denser geographies not being reasonable in that they could viably be built in those geographies, as further addressed in Part 3. 

Q7:  How much of a role can rental MH play in increasing affordable housing, especially recognizing that renters have, on average, lower income than homeowners and thus are deserving of priority policymaker attention?

A:  Unfortunately, there is little federal-level policy focus on MH rentals.

While conducting research for this paper, I did not find any articles dedicated to pure MH rentals, although they probably exist with low visibility. It is a significant lapse and reflects how much the Washington housing policy community is mainly focused on homeownership. This is despite the fact that about 35 percent of American families have historically been renters, including the least economically advantaged who most deserve the focus on affordable housing.  Rental MH is definitely an area worthy of research and policy development.

Q8:  How much do MHCs – specifically those that utilize the own-rent business model – create undue household instability, particularly when “mom and pop” community owners sell out to institutional investors?   

A:  Residents of MHCs that operate under the own-rent business model, which excludes resident-owned communities (ROCs) and pure rental operations, are naturally subject to potential landlord-generated instability, including the extreme case of forced relocation.  The current trend of small and local community owners selling out to institutional investors does appear to increase that instability risk. 

MHCs operating under the own-rent business model put household instability at risk just like any other MH own-rent situation, as the landlord can potentially raise the rent of the underlying land at an undue pace.  In addition, in the own-rent configuration, there is the unusually destabilizing and potentially devastating case of forced removal, and this includes MHCs.  The policy focus on institutional investors, who have been active in buying MHCs in recent years, stems from how most such companies seem to operate under a business model in which they invest significant dollars to upgrade older (and often run-down) MHCs, but then raise rents to provide a return on those investments.  Even when the new MHC owner charges fair lot rents for the upgraded investment in the community, the result can be more than many existing residents can reasonably afford. 

This is one more example of the weakness of the own-rent MH configuration, which as discussed above I consider to be only marginally homeownership at all.  Certain MH supporters are proposing that the solution is to require that all MHC sales be structured to allow a period of time for non-profit organizations (in particular community land trusts) to purchase the community.  There is also the long-standing alternative of prioritizing ROCs, where the residents become in control of their destiny with no third parties involved.  Some MH supporters even indicate a wish to eliminate any for-profit ownership of MHCs.  In Part 3, this challenging situation will be addressed. 

Setting the Stage for Policy Development

To sum up, in my review of articles and videos about MH, I encountered a consistent theme from MH supporters:  that MH is a relatively simple market that can provide much more affordable housing if only two straightforward things were done:  get the government mortgage agencies to prioritize supporting structure-only loans and have state governments eliminate specific MH restrictions in local zoning. 

My analysis of the eight key policy topics above reveals that quite the opposite is true:  the MH market is incredibly varied and complex, primarily but not exclusively because of the frequent separation of ownership of the underlying land vs. the structure. This directly leads to my recommendations for MH to provide more affordable housing not being simple either.  In Part 3 of this series, I will develop and describe policy solutions and approaches consisting of a range of targeted recommendations that reflect the complexity of the MH market.


[1] Freddie Mac, Fannie Mae, the Federal Housing Administration (FHA) and the Department of Veterans Affairs (VA).

[2] Support from the four government mortgage agencies also carries with it taxpayer subsidies, although they are often covert and indirect, and so hard to value. 

[3] The reality is that the range of price points and quality is wider for site-built homes, as they go from very modest to extremely expensive. By comparison, MH structures have a relatively narrow range centered on a low-to-modest price point. 

[4] The three levels of quality are described in a 2017 article by Mobile Home Living (

[5]In the Mobile Home Living articled cited above, it says “If you drive by a manufactured home dealership in the South, you will likely see signs advertising single-wides for $19,995 and double-wides for $39,995.  These are your most affordable models.  They are really nice homes but there is a reason they are so affordable. These models use budget materials and may have staples and glue in lieu of nails.  The flooring will be thinner and the cabinets made of cheaper material.”  Obviously, such price points (which are usually only found in warmer climates, as cold weather requires more substantial materials) cannot support the quality of materials or construction found in site-built homes, although they do create extremely affordable options and help reduce the average cost cited by the MH industry.

[6]In a similar vein, the typical site-built home today is not 100 percent purely site-built anymore.  Instead, over time, more factory-produced and truck-delivered components are being utilized in order to reduce costs, e.g., pre-made roof trusses. 

[7]Cited in “What is the Life Expectancy of a Manufactured Home?” published by Tyrone Woods Company (an MH community).  The Manufactured Housing Institute, an industry organization, claims it is 56 years. Not surprisingly, their number is somewhat more favorable than HUD’s. 

[8] See Fannie Mae’s “Manufactured Housing Landscape 2020.”

[9] This is not to deny that the quality and durability of MH structures have improved tremendously over the past 50 years (as have site-built homes to some degree). However, across the entire range of price points, it has on average less expensive structural elements, along with the negative of mostly not being attached to a traditional, stronger foundation.  This, however, helps preserve its affordability, a topic addressed further in Part 3.

[10] See the article from September 2018 -

[11] It was considered to be high-end because (1) the MH units in question qualified as real property, a term defined specifically below, and (2) they had been financed by the two government-sponsored enterprises, or GSEs, which have conservative requirements as to which MH units they will finance.

[12] See “2022 Manufactured Housing Facts,” where the average cost per square foot is cited as being $72.21 for MH versus $143.83 for site-built housing; the ratio of the former to the latter is almost exactly 50%.  The calculation is based on government data. 

[13] One estimate is the average cost to transport and site an MH structure, which is not part of the factory-cost calculation but is needed to make the comparison more apples-to-apples.  Another estimate is to subtract from site-built homes the cost of the land to get a pure structure-only cost. 

[14] Such a favorable bias has already been seen in the MHI’s estimate of useful life being noticeably longer than HUD’s. 

[15] See “New evidence shows manufactured homes appreciate as well as site-built homes.” Page 2.

[16] In Part 1, because of this, residential structures are designated as generally long-term (GLT) depreciating assets and land a GLT appreciating asset. 

[17] “What Are the Real Costs of Purchasing a Manufactured Home?” from, October 2022, page 9.  (“55” refers to people aged 55 and older.)

[18] See the article from September 2018.

[19] See the Washington Post article from September 9, 2021, “A Factory-Built Home as a Means to Affordable Housing.”  Also note that the picture at the top of this article is indicative of a more typical, recent MH single-wide unit.

[20] According to the Federal Housing Finance Agency, the average home price increased by 44 percent from 2019 Q4 to 2022 Q2 (not seasonally adjusted, purchase-only transactions).

[21] There are also many variations on the three configurations.  For example, sometimes an MHC can have the land all owned by a single landlord company; sometimes it can have each lot individually owned by the structure owner; sometimes it can be owned by all its residents on a cooperative basis; and sometimes it can all be owned by some of the residents with the rest renting from them.  In fact, ROCs have features of both the own-own MH and of the own-rent MH. 

[22] It does not matter if the resident rents both the structure and land from the same owner, or two different owners.

[23] There is an unusual variation on own-rent MH where there is zero rent paid, for one of various possible reasons (e.g., being located on a family farm and occupied by a family member); this would be called “own-zero rent MH.”

[24] In my research, I found that the claimed homeownership benefit in the own-rent case is that the resident family can physically alter the structure without landlord permission – such as paint the walls, rebuild a closet, etc. Again, this appears to me to deliver only very marginally on the usual benefits of homeownership.

[25] See “Challenges to Obtaining Manufactured Home Financing,” by Goodman and Ganesh, Urban Institute, June 2018.  I note that they calculate two spreads, one near 3 percent and one about 4.4 percent, which I have combined above to be “about 4 percent.”

[26] I have read several proposals calling for state legislation to be passed to designate MH structures as “real property,” with the implication that this would enable them to be financed with conventional mortgages.  That is incorrect.  Regardless of the legal label, having only the GLT depreciating structure as collateral, rather than both the structure and the land, is an economic reality that simply cannot support the traditional 30-year mortgage rate and terms. 

[27] For example, having a 30-year maturity, being able to freely prepay at any time, and being able to lock up the interest rate up to three months ahead of time. 

[28]The phrase “to-be-announced” comes from the fact that one feature of the agency MBS market is that the exact securities to be delivered from a seller to a buyer are chosen just before delivery, rather than at the earlier time of the actual agreement to sell/buy.  This depends on the very high degree of standardization found in that market. 

[29] This should not be underestimated.  The liquidity of other bond markets – like corporate bonds – are nowhere near as great.  Only the agency MBS market comes anywhere close to Treasury-like liquidity.

[30] And it is not available to all first-lien home mortgages.  Various standardization requirements preclude a portion of the first-lien market from participating as well.   To start with, only loans guaranteed by the four government mortgage agencies are allowed. There are also more detailed standardization requirements, e.g., a maturity longer than 30 years is not permitted.

[31] This is done through an industry organization, the Securities Industry and Financial Markets Association (SIFMA).

[32] They do such custom securitizations for multi-family (i.e., apartment house) mortgages.  Also, some might think the government agencies could purchase, on a policy-driven subsidized basis, various types of non-securitized loans and that MH structure-only loans could be among those. However, this is incorrect as, especially since 2008, government legal agreements, regulations, and rules in practice limit the agencies to being solely channels for securitization.

[33] This is a fundamental design challenge for the government mortgage agencies in all their activities.  The theory is that competition among primary market lenders would push virtually all the benefit through to actual borrowers.  For the traditional single-family mortgage market, where there are more than 1,000 such primary lenders, it often happens fairly efficiently, although not in all market conditions.  For the structure-only lending market, there is much greater concentration among a small number of lenders – the top five firms account for about one-third of the total market, for example (according to the Urban Institute, as cited in “Four Ways Financing Differs for Manufactured Homes,” July 2018,, thus making the assumption that lower costs to them would be heavily passed through to borrowers much more questionable.

[34] MH (and modular) units have to be transported from the factory to the lot upon which they are sited.  That can vary tremendously from being nearby and thus low cost to being very far away and high cost.  It is not obvious that this cost, on average, is lower or higher for lower-density geographies than it is for higher-density ones. 

[35] The high cost of land, along with the costs associated with local requirements and building fees, is one of the major factors in there being a shortage of affordable (i.e., low-price-point) new housing construction, as developers simply cannot get an acceptable market return on their investment without building more expensive housing. 

[36] The industry lobbying is based on claims that MH units – especially double- or triple-wides – are just as good as site-built housing.   Unfortunately, and I would argue for good reasons already discussed above, that view of MH is not broadly held among the general population, which still regards MH as a second-rate source of housing.

[37] Where MH is not subject to local building codes, but instead to HUD’S national set of rules, modular housing must meet all local building codes.  This raises the cost of modular housing, which also enjoys the efficiency of factory production. 

[38] “Single-family MH” refers to an MH structure sited on a traditional, stand-alone residential lot to provide a typical single-family residence that happens to have an MH-built structure.   This is in contrast to an MHC.

[39] For example, the lower incomes of MH residents will tend to result in less spending to maintain a property than found for the typical site-built home.

Donald H. Layton

Donald H. Layton is a Senior Visiting Fellow from Practice. Prior to joining the NYU Furman Center, he served as a Senior Industry Fellow at Harvard’s Joint Center for Housing Studies, where he wrote extensively about the Government Sponsored Enterprises (GSE) of Freddie Mac and Fannie Mae and more broadly on housing finance. Before his stint in academia, Layton was the CEO of Freddie Mac from May 2012 until June 2019, where he championed the development of Credit Risk Transfers, one of the most significant reforms to the housing finance system in decades.

« Previous | The Stoop | Next »