Unmet Demand Studies
The key components of a Taxi Unmet Demand study (also often referred to as a Hackney Carriage Unmet Demand survey) are the measurement of patent unmet demand and latent unmet demand.

The level of demand is assessed by means of surveys.  These surveys encompass surveys of usage and waiting times at taxi ranks and surveys of the travelling public.  

Surveys of the public will be used to assess perceived waiting times for taxis and any latent demand which may be present along with gathering data on perception of service levels,          
accessibility of services for disabled people, availability and effectiveness of existing taxi ranks and desire for additional taxi ranks.

Surveys at taxi ranks are undertaken to assess the level of patent unmet demand.

DfT guidance on Unmet Demand surveys
In 2006 the Department for Transport (DfT) issued Good Practice Guidance to licensing authorities on the Hackney Carriage and Private Hire Vehicle (PHV) industry. This guidance was updated in 2010. It includes advice on the measurement of unmet demand.
Significant Unmet Demand (SUD) has two components:
Observed or 'patent' demand - that which is directly observable
Latent or 'suppressed' demand - that which is released by additional supply.
Where a limit to the number of Hackney Carriage licenses has been imposed, the DfT recommend that surveys be repeated
every three years to confirm that unmet demand had not arisen.

Observed unmet demand
This is determined from direct observation of passenger waiting times at
representative taxi ranks and at representative times of day. Where the supply of taxis at a particular time and location is inadequate, intending passengers will have to wait until a taxi arrives. Where this waiting time becomes excessive there is unmet demand and where this occurs at a number of locations and for lengthy periods it constitutes Significant Unmet Demand.

Latent unmet demand
Where potential passengers are deterred from using taxis through the assumption or knowledge that waiting times will be high, these passengers may decide not to travel or use an alternative means of transport. These passengers will not feature in the taxi rank surveys.
Therefore to get an estimate of this latent demand an alternative form of survey is required. This generally consists of face to face interviews with pedestrians to enquire about their experience in hiring and using taxis. Such a survey can also provide other information on taxi use.

Other Surveys
The DfT guidance also recommends that stakeholders such as taxi providers and representatives of groups which rely heavily on taxis are contacted for their opinions on the number of taxis and the possible impact of licence quantity controls.

Taxi rank surveys
It is important to capture normal behaviour during taxi rank surveys.  The presence of highly visible survey staff at the taxi rank for several days can affect the behaviour of Hackney Carriage drivers and indeed the behaviour of passengers.  In order to minimise the impact that undertaking the surveys will have on behaviour, we recommend that the surveys be undertaken by video survey.  Video cameras can be fixed temporarily to street furniture such as lamp posts and used to record activity at each of the taxi ranks.  The recordings will be processed to measure waiting times for passengers using the ranks and waiting times by taxis using the ranks.  

Data for week day and weekend activity is required.  It is important that both busy and quiet times are observed during the surveys.  It is often the case that incidences of passengers waiting at taxi ranks can occur at periods of low demand in mid morning or mid afternoon.  Whilst isolated incidences of unmet demand at these times is generally considered to be non-significant, a failure to take observations at these times can form a basis for challenge.

Weekend activity is commonly felt to occur from Friday evening through to Saturday night.  Sunday is seen as a separate and distinct behavioural day.  

The use of video recordings to undertake the surveys offers further key advantages:

The recordings can be used for quality assurance checks
Recordings offer the option of replaying periods of intensive activity, to ensure that passenger and taxi activity is recorded accurately
In order to reduce survey costs during quiet periods, video records offer the opportunity to play the recordings at faster than real time during quiet periods, in order to minimise the cost of data processing whilst providing more comprehensive temporal coverage.  This is especially useful for locations where evidence of little or not activity is required.
Surveys conducted manually are generally selective regarding the periods covered by surveys and not all operating hours are covered for all ranks.  This video recording approach offers more comprehensive, robust and defensible coverage at significantly lower cost.
Video cameras can be deployed to record multiple taxi ranks simultaneously.  This reduces the potential for modified behaviour and will record ‘normal’ activity at the ranks.
The video cameras are installed in small relatively unobtrusive boxes and do not generally cause drivers or passengers to modify their behaviour
Video surveys can be undertaken in advance of the stakeholder consultation in order to minimise the risk that drivers 'get wind' of the surveys and modify behaviour
The opportunity to implement quality assurance checks with manual surveys is limited and expensive and hence, often excluded from the process.
Video recordings can be used to demonstrate features at some taxi ranks such as poor queue discipline

With respect to quality assurance, with recorded activity, it is possible to replay specific periods and undertake spot checks on the activity data at the rank.  

During the most active periods at taxi ranks, there can be a lot of activity with different groups of passengers arriving and potentially poor queuing discipline, coupled with groups which arrive as a single party but hire multiple Hackney Carriages, or indeed groups where not all members actually end up getting in a Hackney Carriage.  Typically this can characterise late night activity when pubs and clubs are closing.  With manual surveys with surveyors present on site, it can be difficult to follow and record all that is happening.  This leads to inaccuracies and estimates at the key times when passenger queuing occurs and unmet demand needs to be most accurately recorded.  In addition, there is no means of checking the accuracy of the survey data.  However, with video, the action can be replayed repeatedly to ensure that the activity is recorded accurately.

If we were to seek evidence that a facility such as a taxi rank is not used for Hackney Carriage activity, video coverage could be processed more cost effectively and more robustly than by posting a manual survey team to record activity.  Four days of recordings could be processed in a matter of a few hours.  

The DfT guidance makes specific mention that any evident delays during peak periods such as evening rush hour or pub closing times should be taken into account when assessing survey data for evidence of unmet demand.  Therefore, it is important that the times surveyed should include a sample of these periods.  However, contractors have, in the past, interpreted this requirement as only busy periods need to be surveyed.  The presence of unmet demand at any time, including quiet weekday periods, should be taken into account.  Therefore it is essential that these periods are also covered.  It is proposed that each taxi rank will be surveyed from 07.00 am to 04.00 am.  

Discreet video camera fixed to traffic signal column.