Thursday, June 30, 2016

How to set on-street parking fees: eleven ways compared

Thursday, June 30, 2016

Many cities poison the politics of parking fees by lacking a clearly-stated basis for rate decisions or by using a wrong-headed approach.

I see this daily in news about parking price reviews from all over the world. Examples from the last few days include Washington DC, Delhi, Singapore, and Richmond VA.

But how SHOULD cities set their street parking prices? By what criteria? And which approaches best serve parking success without parking excess?

This post takes a look at eleven different ways to set on-street parking rates.

The best approaches make improved parking-management outcomes their primary goal. Losing sight of this goal in fee adjustments is a big mistake. If the public can't see any visible benefit from parking fees they will naturally be suspicious of any increase.

Singapore is finding this out at the moment. Cost recovery was the main justification given for the just-released price increases in government-owned parking (HDB and URA parking, which includes paid on-street parking). There was almost no mention of the parking management role of parking fees. Outrage on social media has been the unsurprising response.

Eleven approaches to setting on-street parking prices 

The information below is based on Table 13 in Section 5.5 of my "On-Street Parking Management: An International Toolkit" published by GIZ-SUTP.  If you see errors or omissions, please leave a comment.

The most promising options are near the end, so scroll down to Option 11 if you want to go straight to my recommendation!

By the way, although this post is about on-street parking, many of the comments below apply to price-setting for government-owned off-street parking too.

1. Political judgement  (with focus on revenue in price debates)

If prices are not adjusted regularly, then their parking management benefits erode gradually. After some time, the management goal of setting fees gets forgotten. Parking fees come to be seen as a ‘tax’ with revenue as their sole purpose. By that stage, even modest rises become wildly unpopular and can have a high political cost.

Examples: Very many cities worldwide, sadly, including Indonesian cities and most USA cities. In 2011 Boston, USA, raised the on-street parking price for the first time in 25 years, from $1/hour to 1.25, justifying it purely in terms of revenue.

Strengths: none

Weaknesses: Ad hoc. Hampers efforts to gain support for pricing as a parking management tool. Talking about revenue goals is futile because revenue as a key objective guarantees public hostility to price rises. Efforts to avoid political backlash by keeping prices stagnant usually fail anyway, since parking management comes to be seen as a failure and as merely a tax, so pricing becomes ever more unpopular. Singapore's current parking fee controversy comes to mind here.

In Singapore payment for street parking is via coupons like this.

2. Fuel price benchmark

Involves linking parking fees to fuel prices.  This seems to be an ill-conceived populist policy aimed at restraining local governments from raising prices ‘too much’.

Examples:  European national rules sometimes link parking fees to fuel prices (for a maximum price). Hungary is an example.

Strengths:  None that I can think of.

Weaknesses: Arbitrary. Not linked with management objectives. Encourages belief that parking must not be too expensive. Provides no guidance for where/when to price and at what levels.

3. Traffic speeds/ congestion (to serve traffic reduction goals)

Implies setting on-street parking prices higher in areas that are the key destinations of congested traffic flows. Also implies setting peak-time parking prices to match traffic peak times.

Examples:   Delhi briefly proposed higher parking fees during traffic peak-hours.  A Bangalore proposal suggested zones based on traffic conditions.

Strengths:   May help complement other demand-management policies to limit traffic congestion.

Weaknesses:  Untested. Ignores parking saturation issues. Difficult to implement as a clear criterion (rather than a goal). On-street parking and its prices are NOT closely linked with metropolitan traffic flows. If on-street prices rise automatically if traffic speed targets are not met, they might keep rising without notable impact on regional traffic speeds.

4. Turnover (short durations)

Aims to ensure that convenient on-street parking spaces ‘turn over’ frequently (usually so that they will be used by shoppers and not for all-day parking). In other words, the aim is short parking durations. This approach often goes together with parking time limits.

Examples: Many cities.

Strengths: Keeping on-street durations short is especially relevant for retail areas, serving the interests of retailers.

Weaknesses: Not clear what level of turnover to target. Even if a turnover criterion is met, saturated parking can still cause problems. Turnover data can be misleading in some circumstances. Turnover is not easily/cheaply measured without digital pricing mechanisms.

5. Public transport fares as benchmark

Involves linking parking fees to the price of a primary alternative to driving, public transport. If public transport fares rise, so do downtown parking fees, avoiding an increase in the attractiveness of driving to the city centre.

Examples:  European local governments often link city-centre parking fees to the cost of a transit ticket (usually to set a minimum price). Budapest is an example.

Strengths: May slightly deter future populist attempts to lower city-centre parking prices. Prevents public transport fare increases from encouraging car use for travel to downtown.

Weaknesses:  Arbitrary – not obvious how parking prices should compare with public transport. May not prevent saturation. Suitable primarily for city centre parking. Defines only a minimum price. No guidance for where to price or on pricing hours.

6. Intensity of development

Parking price zones (usually concentric) typically match boundaries in the intensity or type of urban development. These boundaries often also reflect past step-by-step extensions of parking management.

Examples:  Seoul and various cities in Europe and China.

Strengths:  Serves traffic mitigation goals and the need to ration on-street parking more intensively in the busiest areas. Tends to match local expectations of where different policies should apply.

Weaknesses: Decisions on zone boundaries and on prices are rather arbitrary. No guidance for pricing hours.

7. Land values

Base parking prices on some percentage of nearby average land values. Based on the idea that parking pressure is highest in areas of highest land values/rents. Also alludes to the idea that parking should pay its share of land rent.

Examples: Proposed in India's national urban transport strategy. Proposed for Ahmedabad.

Strengths: Simple. Sends a helpful message about the value of parking. Parking pressure may correlate roughly with land values.

Weaknesses:  Untested. Proposals so far make parking an arbitrary (and low) proportion of estimated land prices. Insufficient evidence that on-street parking pressure correlates closely with land rents. Some low-land-price areas may have saturated on-street parking. No guidance for pricing hours.

8. On-street prices higher than off-street

Takes nearby off-street parking prices (or an average of such prices) as a minimum benchmark and sets on-street prices a certain amount or percentage higher. Aims to discourage cruising for on-street parking and encourage use of off-street parking, which otherwise is often under-used.

Examples:  Medellin; Beijing; Frankfurt-am-Main

Strengths:  Probably simple to implement. Widely advocated. Provides market-responsiveness if off-street parking has market-influenced prices. Should encourage off-street parking use and discourage cruising for on-street parking. Is in line with the fact that motorists often value well-located and well-managed on-street parking more than less-convenient off-street options, at least for short-term parking.

Weaknesses:  Limited evidence. Danger it may prompt calls to control off-street prices. Off-street parking is often under-priced due to other policies. Problems if off-street prices are not responsive enough. Does short-term on-street need to be pricier than short-term off-street, or is it adequate if on-street price for 6 hours or more is higher than the daily off-street price aimed at employees?

9. Precise occupancy targeting with tiny zones

Price setting based on a relatively narrow target range for the average on-street parking occupancy (such as 70 to 90%). Frequent price adjustments (monthly for example). Prices can change for any street section and any part of the day in which average occupancy over the previous survey period falls outside the target range.

Examples:  Los Angeles (in Express Park trial areas); San Francisco (in SFPark trial areas).

Strengths:  Highly targeted at a widely supported and important on-street parking management objective (preventing saturation). Tiny zones enable price-sensitive motorists to use their parking location choice to avoid high parking fees. Effective at reducing parking saturation and its ill effects. Makes on-street parking prices very responsive to changing conditions.

Weaknesses: High precision in space and such frequent price adjustments may not be necessary to achieve most goals. Has been tried so far only in high-income cities. Problems conveying detailed price information to motorists. Requires high capacities in data-collection, management analysis and price adjustment.

10. Non-systematic occupancy targeting

An approximation of occupancy targeting emerges if:

  1. avoiding saturation is an important consideration in price setting (with or without an explicit occupancy target range); and 
  2. there is a willingness to have different prices for specific locations or streets with high parking pressure.
Examples: Vancouver, several Boroughs in London, many places in Australia, various cities in Hungary, Taipei (almost systematic actually). Many cities that seem to use another criterion in this table may, in practice, use occupancy in this way behind the scenes.

Strengths: Widespread. Reaps some benefits of occupancy targeting (to the extent that occupancy does influence prices). May provide first steps towards systematic occupancy targeting.

Weaknesses:  Not sufficiently objective or transparent if price setting is a judgement by officials (potentially influenced by other issues besides avoiding saturation). This makes it difficult to defend price changes. So there is a risk of sliding back towards the ‘political judgement’ approach.

11. Occupancy targeting with simple zones

Price setting is based on a target range for the average on-street parking occupancy (or vacancy) rate. Price zones are not tiny (covering several streets or blocks but usually not more than about 1 km across). Price adjustments regular but usually much less often than monthly. Certain examples have some time-of-day pricing but most have a single price for all priced hours.

Examples:  central Auckland; central Calgary; Rotterdam; Seattle; possibly Budapest.

Strengths:  Well targeted at a widely supported and important on-street parking management objective (preventing saturation). Reduces the incidence of on-street parking saturation (and its ill effects such as illegal parking, double parking, and cruising for parking). Makes on-street parking prices responsive to changing local conditions. Simpler price information to motorists than the tiny-zone option above. Does not stretch data management and implementation capacity. Suited to incremental introduction and improvement.

Weaknesses:  Small areas and short periods of severe parking saturation can emerge if parking demand is not uniform within each zone and across the day. [However, this can be addressed by incremental improvements, such as splitting zones or adding time-of-day pricing as needed, as was done in Seattle’s Chinatown.] Zones that are too big often fail to give motorists the option of avoiding high prices by parking a little further then walking.


My current view is that the best choice is occupancy targeting, especially the final option, ‘occupancy targeting with simple zones’.

Most of the other approaches will typically fail to tame on-street parking saturation and its negative side-effects. This risks having pricing being seen as a failure. It also fuels the (often false) perception of parking shortage and prompts various wasteful parking efforts and investments and many misguided parking policy efforts, including excessive minimum parking requirements .

Occupancy that is neither ‘too full’ nor ‘too empty’ is a simple and intuitive criterion that is easily explained to the public.

If you want parking success without parking excess, then urge your city to consider ‘occupancy targeting with simple zones’ for on-street parking price setting.

Do you agree?



  1. Great synthesis, but I would say that the larger issue is what role should cities have in the provision of parking on the public domain? Not how best to price this asset, but rather how best to allocate the use of this space. Street space is precious and has many competing uses from dedicated collective or active transport lanes to better pedestrian environments or parking for shared vehicles, bicycles, etc. I think that the desired goal should be to remove on-street parking and let the private sector worry about how best to charge for it!

  2. I think this is even better than all those :):

    Budget Precise Occupancy Targeting

    Payments data is used in place of sensors to gauge occupancy. Rates are by car length, and payment is encouraged via SMS or smartphone app/website, but with one parking meter per block face, reducing the frequency and cost of collecting cash. Most enforcement is passive, via license plate scanners on police cars that already have them (or maybe a meter maid on a bicycle). No time limits after the initial rollout, after the right prices are discovered. Customers are refunded for time not used, so less likelihood of staying without paying. Scofflaws are automatically sent fines at say, the price of one full day of parking and vice versa - people still paying but not parking are refunded.

    And perhaps finance the improvement through giving the contractor that implements a share of the increase in revenue - but require that his system use an open modular interface.

    1. Excellent point about making payment proportional to car length. I confess that I have not thought enough about it yet. But a recent trip to Germany (where there are numerous little 'smart cars') made me realize I should look into it.

      I guess doing so has been difficult to implement in the past. But it should now be feasible anywhere that uses pay-by-plate, since the system could store the vehicle length and charge accordingly.

      The rest of your points are more about implementation details and mechanisms rather than the price setting itself. But I think I agree with most of these suggestions on the details on on-street fee implementation.

      In fact, cities around the world that have never had parking meters could probably go straight to mobile payments only. Maybe no need for "one parking meter per block face". If I understand correctly, Tel Aviv and Shenzhen for example, already have only mobile payment options for on-street parking. Tel Aviv has several options in direct competition, which also means there is no need for exclusive contracts with payment system providers.

      Thanks for the comment!

  3. What is not accounted for are curb spaces lost to adjacent property owner for curb-cuts to access its _private_ parking space(s). A shared space on the street will be better used than an off-street one -- and the loss of private parking inhibits private car-ownership, another worthy goal.

    Also, consider the long-term situation in which carsharing and self-driving cars reduce the car "population" in different areas of town. Will this drive prices down? Or should there be a plan in place to promote conversion of off-street parking to non-parking uses. Obviously, shared cars are best parking in public spaces, not on private land (much of it in secure underground areas).

    I disagree with "Asher" that on-street parking should be phased out: parked cars are a nice safety buffer between the walking spaces and driving spaces. They are also more efficient in the sense that access lanes are not needed separately, as they double as through driving lanes.

  4. Thanks Zvi, Asher and Chris for your insightful comments!

    The wider issue of how much and what kinds of on-street and public-realm parking to have was not the aim of the post. But I agree it is an important issue.

    And I agree that the design/planning of streets and public spaces should come first, with the design of parking then made to serve street design. Much depends on local conditions, including street widths and local attitudes. In some places, there might be a strong case for eliminating most on-street parking (as Asher suggests). In other places, it might be best to use parking to serve various street-design goals (some of which Chris mentions).

    1. Oops. I see it was Zvi, not Asher, who suggested phasing out on-street parking.

  5. Great list. I see two separate issues being addressed.

    One is how to most efficiently manage a space that is designated for parking. For that shoupista style pricing works really well, and from your list I like 11 the best. Calgary has a pretty good implementation (combing on and public off-street into same system).

    The other is the question what is the best use for the space. Is it really parking or is it something else? In other words, what's the opportunity cost of the parking space? That's a very hard question and it depends on many contextual parameters. Your point 8 gives a very simple way to put a value on providing pubic space for parking by comparing it to a nearby off-street space. Parking on street is more convenient for the driver and pretty much worse for everyone else. (I don't by the "buffer" argument mentioned above, a nice nature strip with hedges and trees is a much better buffer.) So that gives a very simple way to set the floor price for parking: Frankfurt sets it at 50% above a nearby off-street space. If on-street parking in some area becomes under-utilized a couple of spots get yanked and put to a different use. Street trees, bike corrals, bike lanes, wider side walks, ... lots of options.

    In my mind, combining 8 with 11 is the best of both worlds. 8 sets the floor price, and 11 decides the price above the floor. If parking is under-used at the 8-price, a couple of spots gets converted to a different use.