Reducing car ownership in Amsterdam
Amsterdam wants to reduce the number of cars, and create more space for pedestrians, cyclists, playing children and green areas. It seems likely that car ownership will decrease as a result of the measures it is taking, but perhaps not very fast.
In this article, which is a background analysis for an article that will appear in OEK, I’ll discuss why and how Amsterdam wants to reduce the number of cars. I’ll also try to assess whether its approach is likely to be effective. See caveats at the end of the article. If you’re only interested in the conclusions, skip directly to the Discussion.
Why reduce car ownership
Many cities want to discourage car use, but reducing car ownership is often a bit of a taboo. This has long been the case in Amsterdam as well. For example, in 2008, Amsterdam’s social-democrat alderman Tjeerd Herrema said it would be ‘an illusion’ to stop the growth of car ownership. Car ownership was seen as a given, not as something one might change.
This changed with the coalition agreement that laid the basis for the new city government in 2018. The agreement announced that Amsterdam would take the lead in new mobility concepts, in order to reduce car ownership and improve accessibility (GroenLinks a.o. 2018). As party leader Rutger Groot Wassink of the Green Party put it: «We want to discourage car ownership in the city and make sure that Amsterdam once again becomes Europe’s cycling capital.»1
A year and a half later, the new city government published its Agenda Autoluw (Gemeente Amsterdam 2019b). The goal of this new policy was to create more space for pedestrians, cyclists, playing children and green areas. «The emphasis is on cars because of their large spatial impact. But that doesn’t mean that we’re anti-car,» social-democrat alderwoman Sharon Dijksma explained.
Dijksma’s point about the spatial impact of cars makes sense. In central parts of Amsterdam, parking and roads for cars take up between 25% and 44% of public land (this includes roads that are used both by cars and bicycles).2 And it’s not really helping that cars are getting bigger.3
Besides space considerations, one could think of other reasons to reduce car ownership. If there are fewer cars, fewer people will use them, which means less polution, safer streets, less congestion and healthier people. Another consideration is the use of energy and raw materials, and the climate impact, associated with car production. Perhaps one could even argue that car-free streets make people happier and reduce crime.4
Of course, some people are more dependent on cars than others. For example, the Agenda Autoluw states that people with a disability must always be able to reach their destination, also if they need to use a car.
Car ownership in Amsterdam
Car ownership in Amsterdam is relatively low by international standards. The number of cars per person has gradually decreased over the past years, but it started to rise again in 2020.
The recent rise might seem to suggest that efforts to reduce car ownership have been unsuccessful. However, it’s probably too early to draw conclusions. For one thing, some measures will take years to show effects. Second, Covid may have played a role in the rise of car ownership after 2020.
The table below summarises Amsterdam’s main policies on car use and car ownership.
|Car use||Car ownership|
|Facilitate||Road network (Hoofdnet Auto)||Minimum parking requirements for new construction; cheap parking permits|
|Discourage||30km/h speed limit; strategic cuts (knip) in road network||Reduce availability of (cheap) parking options|
|Provide alternative||Facilitate cycling and walking; improve public transport||Car sharing|
Since parking regulation appears to be a key tool to influence car ownership, I’ll start with a discussion of this topic. After that, I’ll briefly discuss ‘car sharing’, and I’ll compare Amsterdam to some foreign cities.
Parking regulation: overview
In most parts of Amsterdam, you have to pay to park your car on the street. Residents can apply for a permit for a fee that ranges from about 33 to 600 euros per year, depending on the neighbourhood. This is far below the market value, which means that these permits can be considered an implicit subsidy for car owners.5
It has been estimated that these implicit subsidies amount to €3,200 per car per year in the central part of the city, or half the cost of owning a car net of parking expenses. High-income households disproportionately benefit from these implicit subsidies, since they are much more likely to own a car (De Groote a.o. 2016).
People who move into newly built houses are not eligible for a parking permit. The original rationale seems to have been that new houses are supposed to come with private parking space, but more recently the measure has been presented as a way to keep the city accessible for pedestrians, cyclists, public transport and cars (Gemeente Amsterdam 2017c).
The number of parking permits that can be issued within a specific area is limited by a ‘ceiling’ or cap. Changes in permit ceilings, shown in the chart below, can be used to identify some recent changes in Amsterdam’s parking regulation.
The chart above reflects the following developments:
- In Noord, the ceiling was raised in 2018, when paid parking was expanded in that district;
- As of 1 July 2019, permit ceilings were lowered, reducing the maximum number of permits across the city from about 170 thousand to about 140 thousand. This will be explained below;
- After that date, permit ceilings were further lowered in small steps, reducing the maximum number of permits across the city to about 135 thousand;
- As of 1 July 2022, the maximum number of permits in Nieuw-West was raised, when paid parking was expanded in that district.
Expanding the paid-parking area
Amsterdam has recently decided to expand paid parking in Nieuw-West and Zuidoost. The expansion of paid parking in Noord in 2018 provides an opportunity to explore the effect such a measure might have on car ownership. The chart below shows neighbourhoods in Noord that were affected by the expansion of paid parking.
In some neighbourhoods where paid parking was introduced, car ownership decreased; these are the neighbourhoods below the horizontal dashed line in the chart above. However, all but one6 of these neighbourhoods saw substantial population growth due to new construction. This means that the reduction in car ownership may well have been caused by changes in the composition of the population.
To the left of the chart, along the vertical dashed line, are neighbourhoods that saw no substantial population growth. In most of these neighbourhoods, car ownership didn’t decrease and in fact increased a little after paid parking was introduced.
This may seem counterinituitive, but it makes sense if you consider that paid parking has advantages for residents who own a car. The costs they face will be low because they can apply for a cheap parking permit. Meanwhile, paid parking will make it easier for them to find a parking place in their neighbourhood, because it discourages parking by people from other neighbourhoods.
Note that expanding the paid parking area is still relevant for reducing car ownership, because other measures - discussed below - only work in areas with paid parking.
Raising permit fees
Ostermeijer a.o. (2019) estimate that raising the permit fee in Amsterdam’s city centre from (then) €500 per year to the market value of €3600, would reduce car ownership between 19 and 24 percent.
Amsterdam has recently raised parking permit fees in a number of areas. In the city centre, the fee rose from €574 to €600 per year. It seems unlikely that such a small increase will have a meaningful effect on car ownership.
Reducing the number of permits
De Groote a.o. (2016) argue that waiting lists for parking permits will reduce car ownership. They estimate that car ownership per household will decrease 2 percentage points (which implies a reduction of the number of cars of about 4%) for each year people have to wait for a permit. You have to own a car to apply for a permit, which implies that you’ll face substantial parking costs during the waiting period. As a result, a waiting list de facto raises the price of a parking permit.
The findings by De Groote a.o. suggest that waiting times would have to be quite long to achieve a substantial effect on car ownership. The authors suggest that from a welfare perspective, ‘infinite’ waiting lists would be optimal. Of course, this would amount to phasing out parking permits altogether.
In 2012, the Amsterdam municipality did the opposite of making waiting lists longer: it set a target to shorten waiting lists for parking permits, because at the time it believed that they were bad for the business climate. Within a short period, waiting lists almost disappeared, largely as a result of removing people who shouldn’t be on the list (Gemeente Amsterdam 2017a). In 2013, 6,247 people were on a waiting list for a parking permit; by 2019, this had dropped to 530 (Gemeente Amsterdam 2021a).
After the 2018 election, the new city government reversed this policy goal. As a first step, it lowered the permit ceiling in areas that had unused capacity. As a result, the maximum number of permits across the city dropped from about 170 thousand to about 140 thousand.
Subsequently, the city government announced that it would lower the ceilings in central parts of the city every 6 months between 1 July 2019 and 1 July 2025. This was expected to increase average waiting times by 2 months. This goal suggests Amsterdam is treading carefully and that the effect on car ownership will be modest.
The map below provides information on current fees and waiting lists for parking permits in Amsterdam (note caveats below).
The number of people on waiting lists rose from 530 in 2019 to about 3,000 now. Estimated waiting times are also available, but the municipality warns that these figures may be incorrect, apparently due to a technical problem. With that caveat in mind, one area has an estimated waiting time of 24 months; ninety percent have an estimated waiting time of under 12 months; and more than half have zero waiting time.
People who move into newly built houses are not eligible for a parking permit. Since parking permits amount to an implicit subsidy on car ownership, one might expect fewer cars in new neighbourhoods. However, things are a bit more complicated.
To the right of the chart above are neighbourhoods with a high share of new construction, where residents are not eligible for parking permits. Some of these neighbourhoods have low levels of car ownership, including a few neighbourhoods with student housing. Others on the other hand have high levels of car ownership. Some of those are in rich parts of the city or have a business park.
This suggests that making parking permits unavailable doesn’t always lead to low car ownership. It doesn’t seem that this can be ‘explained away’ by characteristics of new neighbourhoods, such as the size of houses. It seems likely that parking requirements play a role. These norms prescribe the number of private parking places that must be created in new construction projects.
Current parking requirements have been set to accomodate existing levels of car ownership, not to change them (Van de Reijt a.o. 2021). The minimum requirement is 0.3 to 0.6 parking places per house; the maximum requirement 1 per house (exceptions apply). For comparison; the average number of cars per household is 0.5 in Amsterdam.
All in all, it is unlikely that current parking requirements will do much to reduce car ownership. In fact, artificially increasing the supply of parking space will likely lower the price, thus encouraging car ownership (while making new houses more expensive).
However, this could change if parking requirements are lowered. Amsterdam plans to revise its parking requirements policy this year. Meanwhile, the city is planning substantial new construction in a number of neighbourhoods, and in some cases, it has already decided that lower requirements will apply. In Haven-Stad, which will consist of 40 to 70 thousand houses in the western harbour area, a maximum parking requirement of 0.2 will apply. New construction along the northern bank of the IJ will have a requirement of 0.3.
All in all, it seems likely that car ownership will be lower among residents of future new construction. However, it will take years before the results will start to show. Further, while Amsterdam plans to add a substantial number of houses, the people who will live there will still represent only a modest share of Amsterdam’s population, so the effect on city-wide car ownership will be limited.
Amsterdam is promoting ‘car sharing’ as an alternative for private car ownership (the term car sharing has been criticised because most programmes are run by commercial car rental companies). Currently, some three thousand shared cars are available in the city, mostly in central neighbourhoods.
It’s unclear whether the availability of car sharing has an effect on car ownership, and if it does, how large this effect is.7 Amsterdam expects that demand for car sharing will not increase until paid parking has been expanded (Gemeente Amsterdam 2023). This suggests that the availability of car sharing will have little impact on car ownership unless it is combined with stricter parking regulation.
Brudner (2023) has compared parking permit fees in a number of large cities in Europe, North America and Australia. Amsterdam is in the ‘high’ category, along with Paris, Edinburgh and Stockholm. This suggests that parking regulation in Amsterdam is relatively strict by international standards.
However, some cities in Asia are far more ambitious about controlling car ownership - taking measures, it must be said, that probably would not be possible under current Dutch law. For example, Singapore has a ‘zero car ownership growth policy’: you need a permit to own a car and the number of permits is capped by a quota system (ITF 2021). Singapore has 120 cars per 1,000 people (Knowles 2023), compared to 279 in Amsterdam.
In Japan, you cannot buy a car unless you have a certificate to prove that you have a place to park it. More than 95 percent of Japanese streets have no on-street parking, and overnight street parking is illegal. In combination with other measures such as toll roads, mandatory checks and car taxes, this raises the cost of owning and using a car, especially in cities. In Tokyo, there are 0.32 cars per household (Knowles 2023), compared to 0.51 in Amsterdam.
In 2018, Amsterdam announced its intention to reduce car ownership, as it wants to create more space for pedestrians, cyclists, playing children and green areas.
A key characteristic of the measures it is taking, is that they don’t affect the interests of current car owners, who benefit from cheap parking permits. Measures to reduce the number of permits (lowering permit ceilings, making residents of newly built houses ineligible for parking permits) are designed in such a way as to leave most current permit holders unaffected.
The rationale is likely political. Plans to discourage car use and ownership often meet with resistance from car owners (opposition sometimes evaporates after measures have been implemented, cf. Van Wee a.o. 2023). A likely contributing factor is ‘negativity bias’: people whose vested interests are at stake tend to respond stronger than people who gain from policy change (Weaver 1986). This makes it politically difficult to introduce regulation, even if this would create a net benefit for Amsterdammers.
Amsterdam seems to circumvent this problem by making sure current car owners are not affected.8 This appears to be a case of ‘grandfathering rights’, which Brudner (2023) considers to be ‘non-optimal but an effective tool’ to mitigate opposition.
So far, car ownership per person hasn’t decreased in Amsterdam; in fact it has increased somewhat since 2020. However, a number of measures that are being taken could reduce car ownership in the future. Others will likely have limited effect:
- Fees for parking permits are so inexpensive that they can be considered an implicit subsidy on car ownership, amounting to over 3,000 euros per car per year in the city centre, according to studies. In some areas, fees have recently been raised, but the increase is probably too small to have a noticeable effect on car ownership;
- Amsterdam is reducing the maximum number of parking permits. This will result in longer waiting lists for permits, which will likely discourage car ownership. However, the pace at which ceilings are lowered appears somewhat timid;
- Current parking requirements often require a minimum number of parking spaces to be created when new houses are built. This creates an incentive to own a car, while making housing more expensive. Amsterdam plans to revise its parking requirements. If this results in lower requirements, this could discourage car ownership in the future. The city government has already announced that lower parking requirements will apply in new neighbourhoods such as Haven-Stad;
- ‘Car sharing’ is promoted as an alternative to car ownership. However, Amsterdam expects that demand for car sharing will not increase until parking regulation has been tightened.
All in all, it seems likely that car ownership will decrease, mainly as a result of limiting the availability of cheap parking permits. The size of the effect will likely depend on the rate at which parking permit ceilings are lowered, and on how strict the new parking requirements for new construction will be. Note that it will likely take years before the effects of parking requirements start to show.
As usual on this website, this analysis has an exploratory nature and may contain errors.
Counting cars may seem straightforward, but in reality it can be tricky. In 1993, for example, Statistics Netherlands (CBS) estimated that there were 27,500 passenger cars in Amsterdam’s inner city, but the municipality claimed there were only 18,000. CBS data was based on the national number plate registration, whereas Amsterdam counted the number of cars parked on the streets at night and added the number of available places in car parks. Differences in the way in which company cars are counted could explain some of the difference, but not all.
Company cars can distort data, especially at the neighbourhood level. For example, during recent years, the number of cars registered in Nes e.o. and Westerdokseiland increased from about 1,300 to 7,100, without major population growth. It seems likely that this has something to do with company cars registered in these neighbourhoods.
(International) comparisons can be especially tricky, because the methods used may vary. For example: what types of cars are included; how are they counted; are they presented per person, per adult person or per household.
By 1 July 2023, the municipal website indicated estimated waiting times for parking permits of up to 286 months. Later that month, a disclaimer was added: «Currently, the estimated waiting time in the address checker is not always correct. We’re working on a solution.» In two steps, estimated waiting times were revised to more realistic values, with a maximum of 24 months. These estimates have since remained unchanged. The disclaimer is still on the website.
In the article above, I quote a number of studies that estimate the effect of specific measures on car ownership. While I doubt that it’s possible to precisely predict the effect of a measure, I think these estimates are useful to get a general idea of the likely size of the effect.
Brudner, Amir (2023). On the management of residential on-street parking: Policies and repercussions. Transport Policy 138: 94-107.
Gemeente Amsterdam (2017a). Voortgangsrapportage Parkeerplan Amsterdam.
Gemeente Amsterdam (2017b). Parkeermaatregelen Amsterdam-Noord 2018: Voor een toekomstbestendig bereikbaar stadsdeel.
Gemeente Amsterdam (2017c). Nota parkeernormen auto.
Gemeente Amsterdam (2019a). Bijlagen Amsterdamse thermometer van de openbare ruimte.
Gemeente Amsterdam (2019b). Amsterdam maakt ruimte: Agenda Amsterdam Autoluw.
Gemeente Amsterdam (2021a). Bijlagen: Amsterdamse thermometer van de bereikbaarheid.
Gemeente Amsterdam (2021b). Omgevingsvisie Amsterdam 2050: Een menselijke metropool.
Gemeente Amsterdam (2023). Nota deelvervoer 2023.
Gemeente Leiden (2020). Ruimte voor een groen, vitaal Leiden: Agenda Autoluwe Binnenstad.
Goodman, Anna, and Rachel Aldred (2021). The Impact of Introducing a Low Traffic Neighbourhood on Street Crime, in Waltham Forest, London. Findings, February.
Goodman, Anna, Scott Urban, and Rachel Aldred (2020). The Impact of Low Traffic Neighbourhoods and Other Active Travel Interventions on Vehicle Ownership: Findings from the Outer London Mini-Holland Programme. Findings, December.
GroenLinks, D66, SP and PvdA (2018). Een nieuwe lente en een nieuw geluid.
Groote, Jesper de; Jos van Ommeren and Hans R.A. Koster (2016). Car ownership and residential parking subsidies: Evidence from Amsterdam. Economics of Transportation 6: 25-37.
ITF (2021). Reversing Car Dependency: Summary and Conclusions, ITF Roundtable Reports, No. 181, OECD Publishing, Paris.
Knowles, Daniel (2023). Carmageddon: How cars make life worse and what to do about it. Abrams Press.
Ostermeijer, Francis; Koster, Hans RA; van Ommeren, Jos (2019). Residential parking costs and car ownership. Regional Science and Urban Economics 77: 276-88.
Reijt, Aukje van de, Nico Dogterom en Guido Scheerder (2021). Evaluatie nota’s parkeernormen gemeente Amsterdam. Goudappel.
Russo, Antonio, Jos van Ommeren, and Alexandros Dimitropoulos (2019). The environmental and welfare implications of parking policies. OECD.
Weaver, R. Kent (1986). The politics of blame avoidance. Journal of Public Policy 6(4): 371-98.
Wee, Bert van, Jan Anne Annema and Sander van Barneveld (2023). Controversial policies: Growing support after implementation. A discussion paper. Transport Policy 39: 79-86.
Admittedly, Amsterdam doesn’t actively communicate its goal to reduce car ownership. But see for example its Omgevingsvisie 2050, which states that measures to reduce car ownership and use are a necessary first step towards becoming a walking and cycling city (Gemeente Amsterdam 2021b). Further, one of the goals of Amsterdam’s policy on shared mobility is ‘reducing car dependency and private car ownership’. The city government also states that it actively tries to reduce car ownership in new construction projects (Gemeente Amsterdam 2023). ↩
Calculated with data from Gemeente Amsterdam (2019a). ↩
Goodman a.o. (2020) found that London Low Traffic Neighbourhoods (LTNs), where through traffic by car is prevented, saw a reduction in car ownership. Goodman and Aldred (2021) found that there was a reduction of crime in LTNs, including violence and sexual offences; public order and possession of weapons; criminal damage and arson; burglary; and vehicle crime (only bicycle theft went up). A possible explanation is that LTNs promote walking and cycling and thus lead to more ‘eyes on the street’. Of course, one would like to see more research on this topic before drawing firm conclusions. ↩
The exception is the neighbourhood Bedrijventerrein Hamerstraat (since renamed Hamerstraatkwartier-West), a business park with 0.85 cars per person. It seems plausible that a large share of cars are company cars. ↩
Many studies have tried to estimate the effect of car sharing on car ownership. However, most are based on surveys, which entails methodological problems. For example, car sharing enthusiasts may be more inclined to participate in such surveys, so results may not be representative of the total group of users of these services - let alone future users if car sharing were to be expanded. A few studies have tried to estimate the effect using data on car ownership and availability of car sharing at the local or regional level, but these do not yield consistent results (Kollec 2021, Buscky and Juhász 2022). ↩
For comparison: the Leiden municipality has also decided to lower parking permit ceilings, specifically in the inner city. In a survey among inner city residents, 62% said they were in favour of limiting the number of parking permits for new residents, while 25% opposed this (Gemeente Leiden 2020). Note the document provides no information regarding the representativeness of the survey. ↩