Book review: William Fischel, Zoning Rules! (2 of 2)

Over the holidays, I read William Fischel’s new book, Zoning Rules! It’s an important text for anybody seeking to understand why zoning (and urban planning in general) was invented and proliferated, and how they influence the shape of our cities and societies.

In the first part of my review, I examined Fischel’s analysis of the origins of zoning. In short, it was the homeowners who dunnit: their demands for restrictions on land use led local governments to put them in place. This is an inconvenient truth for critics of planning, as it suggests that there will be ongoing, bottom-up pressure to reverse any liberalisation of controls.

Fischel Zoning Rules cover

This week, I want to look at Fischel’s analysis of the economic efficiency of zoning. It’s well thought out and a bit counterintuitive.

Fischel focuses mainly on “exclusionary zoning”, which refers to policies like minimum lot sizes or minimum dwelling sizes. These policies implicitly require people to have a certain minimum amount of income or wealth to live in a suburb or neighbourhood, as land is often expensive.

A local example is Auckland’s single house zone, which requires a minimum lot size of 600m2. If you want to know why house prices in much of the isthmus start at upwards of $1 million, it’s because you’ve got to buy lots of expensive land to get a house there.

The single house zone in the notified Unitary Plan (Source: Aaron Schiff)

Exclusionary zoning is a hard case to defend, because it is often predicated on keeping “the wrong types of people” out of a place. This can have serious negative consequences across multiple geographic scales:

Zoning is not simply retarding the mobility of the poor from central city to suburbs. According to an important study by Peter Ganong and Daniel Shoag (2013), land use regulation is retarding the mobility of low-income people within the entire nation… [They] show that rising house prices have substantially reduced America’s intranational tradition of moving to opportunity. Like some other researchers, they have found that high housing prices repel potential workers from regions (Raven Saks 2008; Young, Varner, and Massey 2008) and that this pattern is strongly associated with land use regulation (Gyourko, Mayer, and Sinai 2013). [Zoning Rules!, pp. 164-165]

Notwithstanding these issues, Fischel lays out the argument in favour of exclusionary zoning. At least in the US, minimum lot sizes interact with local property taxes (and local expenditures) and with the large number of local governments in the US. In this context, they can play a role in managing fiscal burdens on local governments, albeit with an economic cost.

Fischel begins by observing that local property taxation in the US funds a range of services and local public goods. This includes some things that are locally-provided in NZ, such as transport, parks, and libraries, as well as many services that are provided by central government in NZ. For example, local property taxes provide the overwhelming majority of funding for primary and secondary schools in the US – education consumes around 36% of local government budgets. Police and fire departments are also funded locally, and some local governments run their own welfare systems, provide water and power, etc, etc.

Many of these things can be thought of as “common-pool” goods. If overconsumed, they become rivalrous – i.e. one person’s enjoyment of them reduces someone else’s ability to enjoy them. However, they’re difficult to exclude people from at the point of consumption. (Although certainly not impossible, as toll roads and congestion pricing schemes show.)

Wikipedia types of goods table

Different types of goods (Source: Wikipedia)

In this context, exclusionary zoning can be used to regulate access to common-pool resources – or, at least, to ensure that new entrants to the community have sufficient financial means to pay for their share. As the logic goes, if you have enough money to buy 600m2 of land, you also have enough money to pay sufficient property taxes to put your children through school. This can be described as “fiscal zoning”:

What the present chapter has argued is that the community itself – its elected and appointed officials, more or less responding to established residents – actively shapes and manages the property tax system in a way that would be difficult to do with any other tax base. Local land use regulation constrains the wholesale tax-avoidance behaviour that bedevils most other potential tax bases. Supplemented by revenues from impact fees and negotiated exactions from developers, fiscal zoning makes most development pay its own way. [Zoning Rules!, pp 160-161]

Of course, a key assumption that Fischel is making is that residents have choices about where to live, both at a national level and within individual urban regions. (Or, in economese, there is Tiebout competition between local governments.) This seems like a plausible assumption in the US, where there is a very large number of local governments – perhaps 85,000 in total. For example, here’s a map showing the fragmentation of local government in Massachusetts, which has 50% more inhabitants than New Zealand but roughly four times as many local governments.

In principle, choice between different local governments that offer different planning rules, taxes, and public services should ensure that everyone has a place to go – regardless of exclusionary policies in any one town. Consequently, Fischel takes a moderate view on the good or evil of individual policies. When commenting on Houston (which lacks comprehensive zoning) and Portland (which combines an urban growth boundary with policies to enable infill and intensification), he concludes that:

It is difficult to be too censorious about either city in a nation where most adults have some choice about where to live. [Zoning Rules!, p. 312]

That being said, Tiebout competition seems to work in principle but not in practice. As I discussed last week, bottom-up demands for restrictive planning rules are widely distributed – every community has its share of loud homeowners who will argue in favour of preventing change. Consequently, tight land use restrictions also seem to have proliferated widely. The exceptions, in the US, are places with relatively undemocratic local governments that do not have to respond to the demands of voters:

The South is different from both the North and the West. The South lacks the fragmented local governments that characterize the North, and it uses the voted initiative sparingly and hardly at all on land use matters [unlike the West]… If the demand for zoning is an expression of popular control, why has the South refrained from developing similar mechanisms by breaking up counties into more autonomous municipalities and enabling the voter initiative for land use measures? The answer is the legacy of slavery and racial segregation… [which] undermined the creation of local institutions… [Zoning Rules!, p. 315]

To conclude, it’s worth considering whether Fischel’s arguments about the efficiency of exclusionary zoning apply in New Zealand. In my view, they do not:

  1. Unlike the US, New Zealand has a small and decreasing number of local governments78 in total. Consolidation of borough councils in 1989 and the creation of the unitary Auckland Council in 2010 are likely to have reduced competition between local governments. This is in some respects positive, as it prevents small councils from simply assuming that their neighbours will house the people they don’t want, but it may also imply an increasing need to restrict local governments from pursuing exclusionary planning rules to prevent them from exercising their monopoly powers.
  2. Unlike the US, local governments have much more limited fiscal powers and responsibilities. Crucially, they do not have to fund schools, which are an important common pool resource that is funded locally in the US. As a result, there is no good reason to regulate to keep people out of individual school districts – I’m looking at you, Grammar Zone – as those schools are funded by everyone’s taxes, not just local property taxes.

Consequently, the case for exclusionary zoning to “keep the wrong people out” is likely to be especially weak in New Zealand. That doesn’t mean there’s no case for urban planning at all – after all, other types of rules are often aimed at managing real nuisances and spillovers associated with development.

What do you make of the arguments for and against exclusionary zoning?

44 comments to Book review: William Fischel, Zoning Rules! (2 of 2)

  • Chris R

    I tend to think the cries of ‘heritage!!!’ and ‘character!!!’ mask a drive towards exclusionary zoning. The panic over three-storey terraces and apartments and the distinct lack of panic over already existing three-storey houses are very interesting.

    Edit: The Grammars may not be funded by local bodies, but they’re a rivalrous good in that more students in zone may decrease the quality of education unless the schools can expand, and will certainly decrease the exclusivity of the education, which people seem to value, considering that other schools of equal educational quality don’t give an equal bump to house prices.

  • mfwic

    We used to have a form of Tiebout competition in Auckland where people and businesses could vote with their feet if they didn’t like the Council or its decisions. I have worked for retail applicants who simply gave up on some Council areas and invested in others as a result of planning decisions. Amalgamation removed that competition completely. Now the Council has a monopoly and effectively says take it or leave it. Some take it and some dont bother.

  • Brendon Harre

    I think we should be careful when generalising results from the US housing market to NZ. We take a lot of planning practices etc from Britain too, with the likes of the garden city movement, British Town and Country Act and the way we copied British planning concepts like greenbelts. Also Britain like NZ is one of the most centralised countries in the world, wrt to % of taxes going to central government, so both countries have very little Tiebot competition. Britain like NZ are outlier countries for the size of its biggest city compared to its second tier cities, with the evidence being these second tier cities being too small not the largest city being too big -‘Urban Economics and Urban Policy’ by Cheshire et al (p.40) Housing markets are not naturally efficient markets -they seem to break down at many different levels. So Peter I think Fischel zoning rules argument is valuable information regarding how to improve regulations for the housing market. Even if it is not the whole or complete depiction of the problem.

    • Brendon Harre

      I wonder if we should have a centralised regulatory body that can overrule density, lot sizes, minimum parking requirements, house sizes…. decisions to make the market more responsive. This should be applicable to both developers and local government.

      Developer covenants can be just as restrictive as anything that councils produce. Christchurch for example was exporting houses after the quakes because no new subdivision would allow relocatable houses.

      New housing supply doesn’t seem to be matching new demand wrt the size of new households in NZ, indicating there is a failure in the market. “67 per cent of the increase in demand is for one- and two- bedroom houses, only 12 per cent of the supply is in that sector; 73 per cent of the new supply is in large homes (four or more bedrooms).” -Generation Rent p.140 by the Eaqubs

      • John Mackay

        These seem very relevant points. While bulk and location rules have a direct relationship to the amenity of neighbouring property owners, exclusionary controls like minimum section size and minimum on-site parking are usually very blunt tools. If there is a genuine need for them (such as infrastructure capacity) then it could be better achieved by focussing on average minima rather than absolute.

        I agree that some central direction is desirable, and it doesn’t necessarily need a “centralised regulatory body.” How about just including it with the endless amendments to the RMA?

        • > I agree that some central direction is desirable, and it doesn’t necessarily need a “centralised regulatory body.” How about just including it with the endless amendments to the RMA?

          I’ve been interested in the approach that Japan takes, described in this post. Local governments draw up their zoning maps, but within constrained boundaries. Higher levels of government determine the rural-urban boundary (or “urbanisation promotion zone”, as they call it), in conjunction with transport planning. Within that area, local governments zone, but they can only choose from a limited palette of zones that the national law provides for.

          There are no zones that have minimum lot sizes, no zones that allow only single-unit houses, no residential zones that don’t allow small-scale local services and retail. There are default rules for height planes, setbacks, density limits and so on. Local governments can be more restrictive than that if they really want, but they need to go through a whole process of justification, public consultation, and so on. Whereas the path of least resistance is to let these things stay at the relatively liberal national defaults.

          Japan’s urban policy certainly has other problems, but it’s a model to consider when balancing the power of local governments with the need to address national concerns like affordability.

        • mfwic

          Hi John. People have tried mean rules before but they throw up problems. If it is based on a mean for an area or zone then the first in with applications get approved and the later ones have to be declined. That type of race produces resentment and isn’t perceived as fair particularly by younger people who find all the development potential is used up. If the mean is only on your land then it creates problems for owners of smaller lots. Owners of large areas can make it work but the average and minimum end up the same if you are cutting one lot into two. But means are not as bad ratios for rules. Someone came to me once because they exceeded 5 trips per 100sqm North Shore had as a rule and asked me what they should do. My advice was build a mezzanine floor and not use it!

      • Peter H

        Agree with “demand is for one- and two- bedroom houses”. The way I’m reading census data. It is posable that just under half of the dwellings in Hamilton only had one room being used as a bedroom on census night.

        • Sailor Boy

          To be fair, Hamilton is slightly different to Auckland in that 25% of residents are tertiary students.

          More importantly though. All population growth in the next thirty years will be 1-2 person households. Retirees, young people without children and empty nesters.

      • Frank McRae

        “New housing supply doesn’t seem to be matching new demand wrt the size of new households in NZ, indicating there is a failure in the market.”

        That’s not a failure of the market – its a failure of planning rules. If you have expensive land and rules that prohibit you from using that land for multiple smaller dwellings the rational response is to build large expensive houses.

    • mfwic

      I agree about the taking anything the US does with a grain of salt. My post grad econ lecturers who had done their Phd’s there all advised that you always had to control US data for racism as it is ingrained at every level of society. It was a major driver of suburbanisation there. The policy reaction to racism can be just as distortionary. The whole sub-prime thing was caused by the government removing prudential rules that while not race based might lead to a race based outcome. Not sure that helped anyone.

    • Angus Robertson

      One problem with looking at London and the UK is the relative value of London is much higher. London is a global banking centre of extremely high value, the rest of the UK is just a western country.

    • America is also different in a democratic sense. The US tends to see local governments as sovereign in their own right, with the power to do anything except what’s prohibited by their local, state and federal constitutions. New Zealand’s local government on the other hand, are an arbitrary creation of central government, which can be reorganised at will, as they were in 1876, 1989 and partially in 2010.

      In that sense, we’re not in the same position as them. America has to justify why exclusionary zoning is bad, and convince thousands of petty councils. We can take a step back, and ask what the scope of local government should be in the first place – why should it even have such broad power to make exclusionary zoning rules? In New Zealand, democratic legitimacy comes from the central government (which people actually vote for, one person, one vote, with proportional representation), not local government (which has very low turnout, unaccountable independents with no party affiliation, a ridiculous block-voting system, wildly lopsided districts, and a franchise favouring property owners).

      I’d certainly favour central government moving to limit the power of local government in that regard. At the very least, we should recognise the links between transport planning and land use planning. The government, through the NLTF, spends a lot of money on Auckland’s transport, public and private. We could, say, require local governments to allow at least a certain level of density and commercial activity near rapid transit stops. Even go a step further, and set minimum requirements to allow for growth in each council, ward, and local board area, if they want to get transport funding. And the same needs to go the other way – central government plans education, fire, police, health, and other services, and these need to be taken into account in local planning.

    • Early Commuter

      Except local government in the UK provides schools and welfare services.

      • Jamie Walton

        Yes, administratively, but as I understand it, a significant proportion (most, if not all?) of the funding is distributed to the councils from the central government.

  • Angus Robertson

    A local example is Auckland’s single house zone, which requires a minimum lot size of 600m2. If you want to know why house prices in much of the isthmus start at upwards of $1 million, it’s because you’ve got to buy lots of expensive land to get a house there.

    This is the reason why Auckland is so very slow to build apartments, you have to buy a lot of expensive land.

    BTW – The map shows massive areas of inexpensive land in Auckland (not highlighted), where exclusionary zoning prevents building.

    • Sailor Boy

      But you couldn’t build apartments profitably if land value were too low. The construction cost is fixed and can only be justified on the reduction of land necessary per dwelling (for large scale development) high land prices makes apartments more profitable and economic, not less.

      • Exactly. Higher land price is a prerequisite for more efficient use. But for the market to supply that three other conditions are required:

        1. demand growth; tick
        2. available investment funding, public and/or private; mostly tick
        3. the legal right to build more efficient typologies; much much more restricted

        • Angus Robertson

          1, Yes
          2, There are better places to invest in building than Auckland, we are in competition with these places. Investment here is slower.
          3, Yes

      • Angus Robertson

        Land Value is derived from its agglomeration value based on population and income, you can sell apartments where people are willing to pay the price based on the agglomeration value of living there. Land Cost is a building cost that eats into your profits, it is the significant variable cost to doing construction. Land cost can be raised by exclusionary zoning, agglomeration value cannot.

        Here is a list of agglomeration proxy and land cost for a number of Australasian cities. Auckland has 1.45 million people worth of value and a median land cost of AUD390,000; Sydney – 4.3 million and AUD360,000; Melbourne – 4.3 million and AUD245,000; Perth – 2.0 million and AUD295,000; Brisbane – 2.25 million and AUD217,000; Adelaide – 1.25 million and AUD200,000. Melbourne (with the highest value to cost ratio) has the highest apartment building rate; Auckland is dead last with Adelaide 4x faster to build apartments than us (Adelaide has restrictive apartment requirements – car parks & balconies).

        • Angus Robertson

          * Correction – we are not worse than Adelaide. Auckland has more under construction and Adelaide has more consented.

          We are about even with a city with fewer people than us and a much lower population growth rate.

        • Brendon Harre

          You are onto it Angus. It is cargo cult thinking to believe that you can use restrictive planning to drive up inner city land costs to the equivalent of much bigger or more successful cities and expect to get the same sort of commercial/residential development as those more successful/bigger cities with their naturally higher agglomeration values.

          Brownlee/Treasury tried that in Christchurch’s CBD and all they succeeded in achieving is driving development elsewhere.

      • Angus Roberton

        So – yeah, nah.

        Yeah – you wouldn’t build apartments in Eketahuna, because even though the land cost is low the agglomeration value is also low.

        Nah – you wouldn’t build apartments in Auckland, because the land cost is too high for a middling agglomeration value city.

        • Sailor Boy

          Funny that we have 1000s of apartments under construction then.

          • Angus Robertson

            So slow.

            We have record prices, but haven’t even got close to record construction levels. Aussie blew through all construction records in 2014, Auckland’s property market is seriously retarded.

          • Sailor Boy

            And both Sydney and Melbourne have higher land values than Auckland.

          • Angus Robertson

            Absolutely, cities of 4.3 million people do create massively higher land values than cities of 1.45 million. There are 2,850,000 very good reasons land values in Sydney and Melbourne are higher than Auckland.

            Unfortunately median land costs are much higher in Auckland.

          • Angus you do understand the limits of relying on median numbers? The median of 1 and 19 and of 9 and 11 is 10. Very big difference in practice. Demographia use over-simplification for a reason…..

          • Angus Robertson

            Patrick, let’s make the argument without medians.

            I would credit you with opening my eyes to the agglomeration value a city can offer, as a function of population and wealth. We have a city of 1.45million, Sydney and Melbourne have 4.3million. You must concede that Auckland generates lower agglomeration values. These agglomeration values are worth something. Apartments in Melbourne and Sydney sell for more.

            Also I understand that if you increase the cost of making something, you get less of it.

            Therefore I believe that if we increase the cost in Auckland land we will get less apartments.

          • Angus you constantly repeat that line of reasoning here but it fails the most important test: Apartments are again being built in AKL at a fast and increasing rate, ergo it fails by simple observation.

            Land and dwellings are not the same thing. Deeply intertwined; but the cost/value of each have a complex relationship. Supply of the latter can accelerate when the cost of the former rises, after all that is almost always a result of a rise in demand. An a rise in demand is a necessary condition for a supply response in a market economy. I would like there to be a supply response by the state too, but it seems that is against the current government’s religion… That’d be ok too if only ‘natural’ supply wasn’t also being restricted by pernicious planning regs… And round it goes. I do see high land value as the lesser of the two evils as it does incentivise more efficient and less wasteful use of land. But that does require the legal right to use that land more efficiently…

          • Angus Robertson

            Patrick, If it were as simple as the definition of “fast rate” that had us in disagreement, then there might be a way forward. However our disagreement lies in what constitutes a priority.

            Land and dwellings are different, but the relationship is not complicated. Land is a commodity used in the creation of dwellings, this is why when the demand for dwellings increases the cost of land increases. In a low cost environment small increase in rents will elicit a large increase in supply. In a high cost environment a large increase in rents will elicit a small increase of supply.

            I find it abhorrent that our city is subjecting its poorest to massive rent increases, so that our wealthy enjoy high returns on their land. Auckland is on course for spiralling rent increases through till 2028.

          • Angus Robertson

            Patrick, Apologies there is an easy compromise between our positions, I’ve been reading too much demographia right wing BS these last few days – it has done my head in. We want exactly the same thing , it is just the motivations that differ – you’re an environmentalist and I’m a pragmatic liberal.

            I don’t give a shit about the MUL, its expansion just a means to an end. I want the costs the MUL imposes and the retardation this causes to be eliminated.

            Adelaide offers a tax subsidy on apartments of AUD15k each, if Auckland does the same then retardation stops and we get building much faster. Now since Auckland is imposing a land cost more than twice that of Adelaide, the subsidy in Auckland will need to be in excess of NZD40k.

  • Some random observations:

    → The North Shore appears to be upside-down, most of that low-density zoning is concentrated in the more central areas close to the CBD. Not a clear pro or con, but it looks very odd.

    → I’d be interesting to figure out how many lots in eg. Ponsonby are actually 600 m² or larger right now. A cursory look on the GIS viewer suggests close to 0%.

    → A 600 m² minimum size means any house will sit on $1,000,000 worth of land, which implies the only type of housing that can be developed will be very large houses. Which means:
    • If people get older and their children move out, if they don’t want to maintain that large house and garden, they must leave the community.
    • Their children, who want their own place but can’t afford a $1,500,000 home, must also leave the community.
    • And obviously, as you grow older and can no longer maintain that big garden, you are unfit for the community and shall be banished to a retirement villa.
    That to me looks like a very broad (and sad) definition of “the wrong people”. Big argument against.

    → The school problem is a very real one, if there are more children than capacity in schools, the situation for parents will get very ugly. Even though the process of toddlers growing up to schoolchildren is very predictable, some city councils in Belgium still got caught by surprise. Then again, if families with children are also “the wrong people” then we have nobody left to live in these big houses.

    If you think about this: we are actually, seriously trying to have a rational discussion about down-zoning to a 600 m² minimum lot size 3 km from the town hall. What on earth is going on over here?

    • Roeland. Ponsonby proper, including St Marys and Freemans Bays, are Victorian and generally have great tight sites; sometimes as small as 200m^2. Grey Lynn, Herne Bay etc, are younger, Edwardian (2 of the houses I’ve owned there were built in 1905 for example, and the sites get up to 600m^2. In fact one was originall 1200m^2 but subdivided in the 80s down to a mere 600!

      Inner Parnell is like Ponsonby proper. Devpo is different again, being a sail away from the city, and sites were always bigger there although settled at a similar time.

  • Lindsey

    I think they dumped all the old Heritage Zone areas into the Single House Zone regardless of Lot size. This is the Res 1 zone in the Isthmus and the Res 3 in North Shore. In my area of Kingsland the sites are Res 1 and are about 337 m2. However, you still need about a million bucks to buy the dirt.

    • David B.

      So why have they set such a high minimum of 600m² then? In Devo, as in many other older parts of Auckland, it’s most common for section sizes to be between 450m² and 500m², and I don’t see the logic in setting the minimum at 600m². We might be better of dropping the minimum to 400m² and getting rid of the requirements for off-street car-parking where possible.

  • Lindsey

    The minimum is what you can subdivide down to. It does not apply to existing already established sites. The city fringe car parking area under the PAUP will have parking maxima not minima so my bit of Kingsland which currently requires 2 car parking spaces minimum won’t require any at all.

  • Peter H

    Here the special character zone which includes every housing type one can name, then called a low density area to try justify as special. Looking at koordinates-population-density-by-meshblock/ there’s a pretty loose definition of low.
    Hamiltons’s Proposed District Plan
    Chapter 5 Special Character Zones – p367
    “Claudelands West derives its character largely from period housing providing links with the City’s early settlement, including bungalows, Arts and Crafts houses and villas. The area also contains the ‘sausage style’ apartment blocks that dominated infilling in the 1960s – 70s and detached second infill development units, commercial activities that support the area are located on the fringe of the area. Overall the area is also characterised by its predominately low-density development. Areas of mature vegetation (including street trees) and front yard gardens are also a significant element.”

  • Developers and the council make a mockery of planning rules anyway.

    The proposed development over my back fence includes:

    1) Terraced housing (29 units) on land zoned for single dwellings (8 units).
    2) 1 car park space per unit despite planning rule of 2 per unit, which means cars parked on footpaths.
    3) Roof tops more than 3 metres above height limits.
    4) 15000 cubic metres of fill material to be delivered when planning rules have a maximum of 2000 for such a small area.

    It’s as though the rules are meaningless to the developer.

    • Guy

      Geoff – developers always push the limit to see what they can get away with. Sometimes they succeed – sometimes they fail. It is completely up to the Council what gets approved or not.

      So if you don’t like the way the rules have been stretched, complain to the Council.

  • mfwic

    Thanks for this post Peter. I have been reading more about Tiebout Sorting. You say above it works in principle but not so much in practice. I am not sure it doesn’t work in practice. The fact that we get a lot of rules from the community doesn’t mean it doesn’t work. Sure there are loud people but if others didn’t want the rules they would get involved too. Tiebout’s sorting can still occur. If someone values the theatre they might be more likely to live in the CBD if they value space for a horse then they might go to Whitford and put up with a long commute. I think you can still have efficient outcomes in a bottom up worlds and maybe we are more likely to get efficient outcomes in a bottom up system where different amenities can be provided organically and you can vote with your feet as opposed to a top down “here is the policy we developed for you people- now live with it” approach

Leave a Reply