Do Government Powers of Compulsory Purchase Help or Hinder Urban Development?


How do legal institutions interact with economic outcomes? I think it’s fair to say that legal institutions often don’t register on economists’ radars, mainly because life is simpler when you assume the law is exogenously determined by factors beyond your control. 

While such assumptions are valid for many types of economic analysis, legal institutions are not exogenously determined factors, at least in the long-run. By engaging with legal institutions, economists may be able to identify ways they could be improved. Of course reforms of legal institutions are slow to implement, but they do occasionally happen. Recent amendments to the RMA being a case in point. By engaging with legal institutions, economists can improve our analyses and help to identify possible opportunities for reform = Win-win.

The high-level question I want to answer in this post is whether these powers help or hinder urban development. My interest in these matters was prompted by the Productivity Commission’s recommendation for Auckland to establish an Urban Development Authority (UDA) with powers of compulsory purchase. That is, the UDA should have the power to expropriate privately-held property for the purposes of urban development. I think it’s fair to say the idea of an UDA with compulsory powers of purchase has attracted interest from across the political spectrum, including the Government (source).

In this post I consider why government’s need powers of compulsory purchase and, second, how should such powers be exercised. This post does not present conclusions or recommendations, mainly because I haven’t formed many yet. Instead, it seeks to identify some relevant economic and legal literature as a means of stimulating discussion.

I welcome your comments: Do you think Auckland needs an UDA and, if so, then should the UDA have powers of compulsory purchase? If yes, then what are the potential risks and how would you mitigate them? I’d also be keen to hear about possible alternatives. Enjoy.

Reference: The post that follows draws on an academic paper I am writing about government powers of compulsory purchase. In the interests of avoiding academic self-plagiarism, I should mention that the paper was written first, while this post was written second.


The historical development of most cities has been significantly influenced by government interventions. The Eixample of Barcelona, the boulevards of Paris, and the highways of New York are all noteworthy examples of government interventions that have left large and permanent marks on the affected cities.

The powers of compulsory purchase governments use to appropriate private property in order to undertake these interventions is a perennially controversial topic. One U.S. presidential candidate, for example, recently described powers of compulsory purchase as: “… absolute necessity for a country, for our country. Without it, you wouldn’t have roads, you wouldn’t have hospitals, you wouldn’t have anything.” (Team Fix, 2016). In contrast to this effusive support, Justice O’Connor’s dissenting opinion in Kelo v. City of New London argued that contemporary legal interpretations of the scope of the powers mean “Any property may now be taken for the benefit of another private party … beneficiaries are likely to be those citizens with disproportionate influence and power …” (O’Connor, 2005).

Given these divergent views, there would seem to be value in developing a clear understanding of the economic evidence on which the powers are based, certainly before we expand the number of government institutions that have recourse to these powers. In particular, I think there is a need to understand why governments need powers of compulsory purchase and how they should be exercised. The standard economic justification for the powers is that — in their absence — information asymmetries gives market power to landowners, leading to hold-out problems. This is where landowners refuse to sell so as to capture surplus associated with the government’s investment. By annulling landowners’ market power, compulsory purchase avoids hold-out problems and enables efficient levels of government investment. It’s important to note that private actors seeking to assemble property don’t face the same issues, mainly because they can negotiate anonymously to buy properties via third-parties.

Exercising powers of compulsory purchase is not without economic costs. Perhaps the most notable costs arise from involuntary transfers of property, which may result in welfare losses for affected landowners. To counter this issue, most jurisdictions require governments to demonstrate that the proposed intervention is in the public interest, and to pay adequate compensation to affected landowners. These legal requirements give rise to two additional economic questions: How should we define “adequate compensation” and “public interest”?

Neither definition is as straightforward as it first appears. Most jurisdictions define adequate compensation using market prices. Some scholars argue the value of property to its current owners, however, includes idiosyncratic attributes, such as sentimental attachment, which are not reflected in market prices (Nadler and Diamond, 2008). Other scholars argue that these idiosyncratic attributes are already accounted for in market prices, while noting that the landowner’s loss of autonomy is not (Lee, 2013).

Definitions of public use are similarly complex. In the U.S. the definition of public use has even been broadened to include wider economic impacts arising from private activities. Under this definition, governments have been allowed to use powers of compulsory purchase to appropriate private property, which is then transferred to private third-parties. The aforementioned opinion by Justice O’Connor in Kelo v. City of New London was prompted by precisely such a situation. Even where the resulting asset is publicly-owned, there are many situations where public access is restricted, for example by user-charges and/or regulations, which are on the surface not readily distinguished from private ownership. From an initial survey, the economics literature doesn’t seem to have a good answer to the question of how one defines “public use”.

Finally, the processes involved in determining adequate compensation and demonstrating public interest may generate non-trivial transaction costs. Imagine, for example, the resource cost of the recent Unitary Plan hearings, where many highly-skilled, well-paid people were paid to sit in a room for months if not years. Is it possible that the transaction costs involved in a particular legal process outweigh its benefits? I think so. At the very least, there seems to be a need to understand potential interactions between legal processes and transaction costs. It may, for example, be more efficient to offer more compensation early in the process, so as avoid transaction costs downstream.

Answers to such questions are not a trivial academic exercise. Evidence also suggests that protection of property rights affects the incentives facing governments and citizens alike, and impacts on long-run economic development (Acemoglu and Robinson, 2005). Powers of compulsory purchase may also have implications for social justice, with some studies finding that land owned by low-income households is more likely to be taken (Carpenter and Ross, 2009). Recent experiences in Auckland, whereby wealthy neighbourhoods successfully opposed the proposed Eastern Motorway across Hobson Bay, provide some indication of why such patterns may emerge. In the U.S., scepticism over the use of powers of compulsory purchase has given rise to significant opposition from across the political spectrum.

Initially simple questions are, upon closer inspection, revealed to be anything but. Understanding why governments need powers of compulsory inspection, and how these powers should be exercised, can contribute to more informed policies on Urban Development Authorities.

Of course, this doesn’t mean that I’m opposed to an UDA having powers of compulsory purchase. Personally, I’ve always been attracted to the notion of a “New Zealand Place-Making Authority” (NZPA) to counter-balance the NZTA’s movement focus on movement. I think the potential benefits could be significant. Nonetheless, the powers such an institution should have is an open-question.

Gnawing in the back of my mind is the fact that powers of compulsory purchase have in the past been used to pursue questionable outcomes. The highways of New York promulgated by Robert Moses, for example, were justified under the auspices of slum clearances. One of the great urban thinkers of our time, namely Jane Jacobs, was spurred into action by a government body wielding powers of compulsory purchase. Would ignoring the risk such things happen again be indicative of the “fatal conceit”?


Acemoglu, D. and Johnson, S. (2005). Unbundling Institutions Daron Acemoglu and Simon Johnson. Journal of Political Economy, 113(5):949–995.

Carpenter, D. M. and Ross, J. K. (2009). Testing O’Connor and Thomas: Does the Use of Eminent Domain Target Poor and Minority Communities? Urban Studies, 46(11):2447– 2461.

Nadler, J. and Diamond, S. S. (2008). Eminent Domain and the Psychology of Property Rights: Proposed Use, Subjective Attachment, and Taker Identity. Journal of Empirical Legal Studies, 5(4):713–749.

O’Connor, J. (2005). KELO V. NEW LONDON (04-108) 545 U.S. 469 (2005).

Team Fix (2016). Transcript of the New Hampshire GOP debate, annotated.

Further reading

Blume, L., Rubinfeld, D., and Shapiro, P. (1984). The Taking of Land: When Should Compensation be Paid? The Quarterly Journal of Economics, 99(1):71–92.

Burrows, P. (1991). Compensation for Compulsory Acquisition. Land Economics, 67(1):49– 63.

Kelly, D. B. (2006). The Public Use Requirement in Eminent Domain Law: A Rationale Based on Secret Purchases and Private Influence. Cornell Law Review, 92:1–66.

Miceli, T. J. and Sirmans, C. (2007). The holdout problem, urban sprawl, and eminent domain. Journal of Housing Economics, 16(3- 4):309–319.

Strange, W. C. (1995). Information, Holdouts, and Land Assembly. Journal of Urban Economics, 38:317–332.


32 comments to Do Government Powers of Compulsory Purchase Help or Hinder Urban Development?

  • Thanks Stu.

    Witnessing NZTA driving roughshod over the relatively powerless Panuku in Onehunga, I’d say we have a very long way to go to get place afforded the powers of vehicle movement.

    Good time to raise the discussion.

    • stu donovan

      Assembling property for transport corridors is arguably the most difficult of all; long contiguous corridors of land, where any one hold out compromises whole project.

      Urban development is a slightly more nebulous beast, at least on a city-wide scale. Its less beholden to individual property owners, and so can more easily respond to hold out problem by investing elsewhere.

      So im not sure UDA needs same powers as Nzta, but it may need some powers.

  • Brendon Harre

    The economic problem is site assembly or contiguity. Land frequently needs to be bought in sequence. If a large or continuous site for a particular use is needed and it requires purchases from multiple owners then the secret buyer tactic is still not ideal. In other markets such as for Labour there is no need for in seqence or contiguous purchases of employment. The question I would ask is how does this market characteristic affect price discovery?

    So due to this problem the second best route sometimea is government compulsory purchase but what problems does that cause? Cera and CCDU in Christchurch were effectively a UDA but they have been ineffective at rebuilding the CBD of Christchurch for example.

    • Stu Donovan

      Why is sequencing important? If you have a project, and it requires a certain number of properties, then you need to negotiate with all landowners no?

      I agree, though, that the secret buyer tactic is not ideal. If you can think of alternative policies that would be better let me know!

      • Brendon Harre

        Contiguity is the better term but people are less familiar with it. An example might be a pie factory. Say it needs 1000sqm of floor space. But property sizes in the city are only 250sqm then the pie maker will need 4 properties that are contiguous or in sequence with each other. It is no good assembling a group of 4 properties which have no sequencing relationship with each other. The pie maker does not have the same contiguity or sequencing problem with other factors of production such as ingredients. The apples for instance do not have to come in an oderly sequence from a particular set of trees in a particular orchard.

        • Peter Nunns

          Contiguity is indeed a better term. Sequencing is ambiguous – it could refer to either temporal order (eg a sequence of actions) or spatial order (eg a sequence of books on a shelf).

          Or we could speak in plain English and just say that it’s important that the parcels that one is trying to assemble are next to each other!

          • Brendon Harre

            Thanks Peter. When I have talked to people about the market imperfect problems of contiguity -I just get this blank look. It is difficult always explaining a concept with long sentences -it is good when a concept gets an accepted/known label. I was trying out the sequencing description to see if that helped the explanation but it doesn’t seem to fit that well either…..

            I wrote an article titled “Why land contiguity is causing market failure in New Zealand’s cities” and I think it turned off a lot of potential readers because they were confused about the meaning of “contiguity”.

            Perhaps the better wording would have been the market imperfections caused by site assembly?

      • Brendon Harre

        So the properties do not have to be bought in a sequence. Just that once all the land is bought they have to be in sequence (or contiguous). There is no point to grabbing random pieces of land just because they collectively add up to the desired quantity.

        • Stu Donovan

          Ah thanks for clarifying. And yes contiguity is a problem.

          It’s also the reason why anonymous third-party negotiations are not the ideal work-around for private sector, because talk between neighbours thereby compromises anonymity.

  • Peter Nunns

    Good post, Stu – an important topic to raise.

    In principle, I can see a case for extending compulsory purchase powers to general urban redevelopment projects. Transaction costs and holdout problems can lead to market failure and a sub-optimally small amount of redevelopment.

    In practice, I am skeptical, for three reasons:
    1. Policy alternatives – while compulsory purchase is probably necessary in order to assemble land for network infrastructure projects, it isn’t the only way we can encourage/enable redevelopment. For instance, if we suspect that small lot sizes constrain higher-density redevelopment, why not cut back setback rules, height limits, and parking requirements instead to allow more height on small lots?

    2. Appropriate compensation – people may value their homes more highly than market prices, and failing to accurately address this will lead to welfare losses.

    3. Political realities – even if compulsory purchase powers exist, any sane politician will use them sparingly for fear of backlash. And they will attempt to constrain use by a UDA, even if it has its own board. Which takes us back to point 1 – are there other things we can do that might also work?

    • mfwic

      I have had a lot of experience with your point 2. In NZ the Public Works Act requires compensation on a willing buyer. willing seller basis. That is simply legalised appropriation (lawful theft?) of the current owners existing consumer surplus. If they only valued their land at the market rate then they would already have their property on the market. The assumption assumes their surplus is zero for compensation. No wonder it is so hard for public bodies to buy land sometimes. I met a man in Otahuhu who grew up there and was going to retire there. He had built a huge house that was his dream home. Real Estate agents would say he over-capitalised, but that is only a problem if you intend selling and he didn’t. The difference between his costs and the market was a lower bound of his consumer surplus. The Requiring Authority low balled him and was actually required by policy to low ball him. It was heart breaking stuff. So if the state is going to use their might then owners have to play games to increase the price. First step is getting resource consents for activities you will never build, intensify the activity they have to pay for. Lease it to a third party who can claim business disturbance. A don’t forget to claim Solatium. Best of all, show that land for the activity isn’t available on the open market so they have to find you a site.

      • Peter Nunns

        As the legal adage goes, hard cases make bad law. We shouldn’t get overly focused on cases like the one you cite, as they may cause us to overlook larger positive effects.

        To me, it raises the question: Can we distinguish between strategic holdouts and high willingness-to-accept values? And how prevalent are these behaviours anyway? The observed outcome – refusal to sell – is the same, but the intentions and underlying motivations may be different.

        I’d like to see more empirical evidence on these questions.

      • Dgd

        I wonder of anyone in NZ has succeeded in increasing purchase price using solatium?
        Nevertheless some good and interesting advice you give, sort of has a Trumpic feel of ‘playing the tax system’ about it.

    • Brendon Harre

      I would agree with Peter in that -yes there is market imperfections in the land supply market to do with site assembly, but compulsory purchase as an alternative option is also flawed/imperfect. So it is not clear what the best option is. For private developments there are alternative examples of voluntary land reallocation and adjustment schemes that have worked in areas with a high degree of trust -Denmark farming communities -Japanese residential intensification……

      • Peter Nunns

        To clarify, I think that the main reason that we currently need large sites to do comprehensive redevelopment is that we have another set of rules that disallows or discourages intensive development on small sites.

        The older areas of many cities include quite a few midrise buildings on relatively narrow sites, built out to the property boundaries on the front and sides. In Japan, they’re constantly redeveloping these sites as buildings depreciate. In many western cities, these areas are legally protected from demolition, but it’s *also* basically illegal to build more of them in incremental fashion!

        This raises various questions of efficiency and logical consistency. One question for advocates of a different policy approach might be: Why is it better to implement new policies to control for the adverse consequences of another existing policy, rather than simply removing the existing policy?

  • Kerry Wood

    Transport generally needs some sort of compulsory powers, going back to medieval river navigations in europe. The often also controlled the tolls that could be charged. c19 Railway construction in Britain was very costly, and lack of adequate powers was part of the problem.

    Yes, Stu, negotiate with all landowners, but recognise that without backup powers they will have you over a barrel. It might run more smoothly if the purchaser had backup powers akin to the Public Works Act, perhaps based on valuation plus a percentage, or plus the reasonable private costs of moving.

    It would also help if the purchaser was a public body and retained some long-term control over the project.

    I am just back from Sydney, where I saw a big cockup at Barangaroo, the old commercial port beside the CBD. The development authority took on cleaning up the old gasworks site (why not the developer?) but specified that a large proportion of the site be public open space. They got it: the most valuable land, close to the CBD, is buildings, and the open space is nicely done but too far from the CBD for lunchtime use, even where meal breaks still exist. Worse, it has no shade. Maybe the contractors could use their rock-shifting machinery between other jobs, but would have needed outside contractors to plant decent-sized trees.
    It is, and looks likely to remain, badly underused: guess what happens next?

    A good residential example is Vauban in Freiburg, with up to seven storeys but also existing buildings and a wide range of house/apartment sizes. Another impressive example, twenty years ago, was an aerial photo of Toronto, with patches of high-rise marking the metro stops.-

    • Stu Donovan

      I tend to agree Kerry – an UDA could have these powers in their back pocket, but be encouraged to negotiate voluntary transactions. Moreover, structure the powers so that it’s in everyone’s interests to settle early. That means internalising transaction costs to affected landowners, i.e. paying them more now to avoid litigation later.

      I also wonder whether an Auckland wide UDA would tend to avoid some of the hold-out issues that affect individual projects. Reason being that the UDA would likely have so many projects on the go that capital constraints were binding. In this case, an intransigent property owner simply means that capital is re-allocated to advance another project.

  • Harriet Gale

    Kelo was a very controversial case even though it decided in favor of New London, the Executive changed its regulations to restrict this on a federal level, and many states have enacted expanded eminent domain protections.

    From the dissent “Any property may now be taken for the benefit of another private party, but the fallout from this decision will not be random. The beneficiaries are likely to be those citizens with disproportionate influence and power in the political process, including large corporations and development firms.”

    Allowing Eminent Domain beyond Public Infrastructure may do some good, but as we have seen before again & again this power will effect poorer neighborhoods over richer, I think this issue should be considered as you have done before we make any decision

  • Kelvin

    There are human rights cases in china, where corrupted local offical use laws and policy to forcefully collect land to government cheaply. And then sell the land back to developers with a huge margin, in order to make a profit and corrupt the money into their own pockets.

    The collection process are injustice and involves violence by offical secretly hired gangs. Police controlled by officials will keep a blind eye. The judges are also controlled by officials.

    I think New Zealand is far better, however New Zealand corruption could be a form of political power laundering, making it legally hard to prosecute. For example a political party selling off land to one of the supporting developers who has close relationship with such party.

  • mfwic

    The problem isn’t just compulsory purchase. It is the Notice of Requirement process as well. It makes agencies lazy. I have worked on numerous structure planning exercises where it would be easy to plan infrastructure from the start but where the agencies take the view that ‘we will sort it out later through a designation’. It means school sites are not located in a sensible place, pump stations get squeezed in and even highways can appear later. Despite all the other stakeholders trying to work together.

    • Stu Donovan

      good point. And questionable NoRs still blight Auckland; Grafton Gully a case in point.

      Tunneled highway to PoA? Yeah right.

  • Matt P

    Interesting discussion.
    I think we need an UDA, as the market by itself will be insufficient.
    But….I’m not a big fan of compulsory acquisition. It’s just got some fairly draconian elements. Maybe if it
    was limited to compulsory acquisition of business land it ‘might’ be OK. But I just don’t like the idea of big brother being able to come in and seize people’s homes.
    Also, as Peter eloquently alludes to, there are other ways of skinning the housing supply cat.
    What sort of UDA projects would typically demand compulsory acquisition? Typically, larger scale, 3-4 storey plus redevelopments.
    With the cost of this form of development, unless it is heavily subsidised, this form of housing will never be affordable. But dense two storey development does have a chance of getting somewhere approaching affordability.
    I’d like to think we could do much nicer looking 21st century versions of the sausage flat. I’ve done designs that can get 6 units on 16m wide by 45m deep sections. 2 beddies at circa 60-65 sq m realised to the market in mid value locations for circa 450-500K.
    These would need the following rule settings to work…. TBC….

  • Matt P

    – No ‘Outlook Space’ rules
    – No private outdoor space requirements*
    – Retain height to boundary rules
    *Instead, I would require things like minimum ceiling heights (2.7m) and minimum glazing / daylighting into units,as well as requiring Juliet balconies on upper levels.
    Private open space requirements are ‘land killers’ which are highly detrimental to yield and affordability.
    But this is just a private fantasy. With the dominance of urban design doctrine in Auckland planning we will never see these things unless there is some totally unexpected huge shift that places housing affordability higher than urban design on the planning radar… .

    • They’ve added minimum internal height rules in Sydney. I agree 2.7m is good for living areas, but not necessary for kitchens and bathrooms, and not only is this space useful for ducts etc, but the living space can be made to feel higher by compressing a kitchen or hall aera for example, so i wouldn’t want to see blanket rules like that without variation or flexibility being allowed.

      • Matt P

        yep I’d be happy with that Patrick
        What are people’s views on doing away with balcony requirements?
        Do we need more public space to compensate?
        how about a trade off where a developer can do away with private open space, but they need to pay a higher reserve contribution to mitigate?
        I say the latter, because based on my time in Europe, lots of apartments don’t have private open space or even balconies, but one of the reasons that can work is they are often near pocket parks etc.
        Another option could be private open space could be done away with if the development site is within say xxxx metres (400?) of a public park / reserve

        • Peter Nunns

          Here’s what the evidence seems to say:

          1. When height limits and site coverage ratios are binding, which they are in most situations, adding a balcony will entail reducing living space in apartments.

          2. This is a worthwhile tradeoff for some people but not others. An analysis of apartment sales suggests that roughly 3/4 of Auckland apartments have balconies. The apartments with balconies tend to be larger than the ones without, and probably (I’m speculating here) higher quality overall. This suggests a segmented market, with large, high-quality apartments with balconies pitched to wealthier people, and smaller, cheaper apartments without balconies pitched to students and lower-income people. Requiring balconies would likely restrict the latter market, which would not be helpful.

          3. Auckland has done a relatively good job ‘future proofing’ its urban parks – we have quite a lot of them. There are relatively few places in the city where people are more than 500-1000m from the nearest park, and many parks are sparsely used at present. So I’m reasonably sanguine about the availability of public open space.

          While I’m at it, if we do want to regulate the internal characteristics of apartments, higher floor-to-ceiling heights are not a totally crazy thing to do. In some respects, the volume (in m3) of small apartments is more important than the floor area (m2) for internal amenity.

          Higher floor-to-ceiling heights are reasonably cheap to deliver *provided* that building height limits are set in a way that allows you to build the same number of storeys. For instance, if your height limit is (say) 16m, requiring a 2.7m floor to ceiling height probably means losing the opportunity to build a fifth storey.

    • kelvin

      Just thinking the possible is to but a large balcony and deck on the roof, with part of the living room, and put the toilets laundry garage and bedrooms down stairs.

      However since airspace underneath of balcony is count as building coverage, that might not be possible.

      The whole thing of building coverage and premutable area need to be revisited for balcony for more efficient use of land

  • SDW

    Is there a case that a UDA would automatically have the powers to carry out CPO’s under the PWA? The definition of public work under the PWA is very broad brush, likewise the purpose of local government in the LGA despite recent amendments.

    • Stu Donovan

      good question. I think it depends on how it is legally structured.

      There may be an argument to structure the UDA as a private organisation, rather than a CCO, so as to 1) reduce liability and 2) remove obligations to comply with LGOIMA (although these can always be refused on grounds of commercial sensitivity). This would also enable the UDA to negotiate anonymously via third-parties.

      In doing so, however, you may remove access to CPO via PWA.

  • John Mackay

    In urban areas, I agree with Peter that the answer is to remove the rules (like minimum parking requirements) that penalise or physically prevent the redevelopment of small lots. But how about rural areas. Doesn’t it make sense on all sorts of levels to allow the ratepayers to reap the value uplift from re-zoning and providing the infrastructure for greenfields development beyond the city limits? It was a commonly used tool in the past, wasn’t it?

    • Stu Donovan

      good point John. The distributional impacts of re-zoning are especially weird in the context of a tax system where (windfall) capital gains aren’t taxed.