APPENDIX 10
Supplementary memorandum by energywatch
Thank you for your letter of 11 October to our
Chair, Ed Gallagher requesting clarification in certain areas
of energywatch's evidence to the Trade and Industry Committee
on fuel disconnections, as well as the written memorandum we provided
on the links between disconnections and other socio-economic and
demographic factors. This is a full response that I trust will
provide all the additional information the Committee needs.
In March 2004, pursuant to the powers conferred
upon energywatch by section 24 of the Utilities Act 2000[16]
we requested—amongst other pieces of information—the
total number of disconnections of domestic consumers for debt
for both gas and electricity for the eight quarters to December
2003 from the major six suppliers, as well as Atlantic Electricity
and Gas. This information was requested by district level postcode
(eg SW1, SE11).
When the information request was sent out in
March 2004 we believed that this level of detail would enable
energywatch to carry out a geographical mapping of disconnection
and, using a piece of market segmentation software called Mosaic,
to superimpose this data onto existing socio-economic information.
The data provided by suppliers varied widely
in terms of detail. The following list describes their information
request returns:
— EDF provided the information by country
only (England, Scotland and Wales), arguing that they did not
collect the information by postcode but that they would in future.
— Powergen provided data only for third
and fourth quarters of 2003 (both electricity and gas), and was
unable to provide earlier Powergen data or data for TXU.
— Scottish and Southern provided data
that met our specifications, but only for 2003.
— Npower provided data that met our
specifications.
— Scottish Power provided data that
met our specifications.
— British Gas (BG) provided data that
went beyond our specifications.
— Atlantic provided data that went
beyond our specifications.
However, once we had begun to analyse the data
we had received it became clear that, in order to have precise
and accurate results, the complete postcode would be required.
Unfortunately, this was too close to the deadline for the submission
of our evidence to the Committee for us to go back to suppliers
and request the additional data for all disconnections we had
received.
We therefore analysed only the data that had
been returned to us with the full post code information (which
was that provided by BG and Atlantic). Since gas disconnections
accounted for 94% of all disconnections in 2002 and 2003, we analysed
only gas disconnections. BG's disconnections represent 74% of
all disconnections for gas. As the information we received from
suppliers only represented 84% of all gas disconnections (due
to the omissions outlined above), the data provided by BG and
Atlantic amounted to 88% of all received data on disconnections.
Therefore, the analysis for gas was done for
a total of 27,476 disconnections for 2002 and 2003, which represented
88% of the available data and 74% of the total number of disconnections
for the years in question (in other words, 74% of the Ofgem figure
of 37,753).
In addition to an analysis of this data with
respect to socio-economic characteristics, as shown in the appendix
to our evidence to the Committee (the table which relates incidence
level of disconnections for each Mosaic type and group), we were
also able to carry out a geographical mapping of this data at
district postcode level. This geographical mapping exercise showed
us the number of disconnections for each area postcode.
We then used the remaining data received from
suppliers which did not contain the complete postcode (that provided
by Npower, Scottish Power, Powergen for the third and fourth quarters
of 2003 and Scottish and Southern for 2003) and compared the distribution
of these, finding that it corresponded significantly to the same
areas as the data we had used for our analysis. We therefore consider
this to be accurately representative of all gas disconnections
across the industry for 2002 and 2003.
As the Committee will be aware, section 24 of
the Utilities Act 2000 allows energywatch to require suppliers
to provide certain information. In practice, we have used more
informal and collaborative methods of getting information from
suppliers rather than the more formal procedure that is open to
us. We also have a "lead office" arrangement, whereby
each of our regional offices has a close, consultative relationship
with one of the six major suppliers (eg our Central regional office
takes the lead in managing our relationship with EDF). We have
generally found this to be an effective and "light-touch"
way of working.
This was in fact the first time we used our
powers under section 24 to request information from suppliers.
After our initial letter was sent out some suppliers requested
additional clarification. We provided this clarification to all
suppliers and formally withdrew a number of our initial requests
in order to allow companies enough time to collect and collate
the information for the requests in which we had a particular
interest.
However, we still found that certain suppliers
complained that our remaining requests were too time-consuming
and would be impossible to fulfil. In one case, when we pointed
out that the company in question only accounted for a very small
proportion of disconnections, it realised that our information
request was not likely to be as time-consuming as it had believed.
Therefore, in general we found that suppliers put up some resistance
but this subsided when they realised that we were requesting information
that should have been readily available to them. Nevertheless,
certain suppliers did not provide all the information requested
and did not feel obliged to do so, claiming that it would be costly,
resource intensive and futile. One or all of these three reasons
were generally given when either insufficient or no information
was provided to us.
We also linked the information we received on
electricity disconnections where we had the complete postcode
to the socio-economic information available to us. However, due
to the fact that BG and Atlantic's data only represented 10% of
all electricity disconnections and 15% of the data received (to
at least district level postcode details), we opted for concentrating
solely on gas, as we felt the relatively small sample size of
241 electricity disconnections with full postcode made the Mosaic
categorisation unreliable and with a potentially significant margin
of error.
In addition, electricity disconnections are
far fewer (just 6% of all disconnections) than those for gas,
leading us to believe that gas disconnections provide a much more
accurate idea of exactly which customers are being disconnected.
The principal reason for this disparity is that it is easier for
suppliers to enter premises and install a electricity prepayment
meter (PPM) than it is for gas, where it is necessary for all
appliances at the premises to be turned off and purged when the
supply is reinstated, which essentially means that the householder
must be present when this takes place.
As mentioned above, the high margin of error
coupled with the singular nature of electricity disconnections
was responsible for the "certain seemingly significant differences"
between gas and electricity disconnections.
Our analysis of disconnections in error was
initially based on 802 complaints to energywatch which were resolved
between April 2003 and March 2004. Of these 280 were rejected
for analysis either because there was insufficient information
or because they did not actually relate to disconnection. The
remaining 522 formed the basis of our evidence to the Committee
and included both written complaints and those taken over the
telephone (the majority of energywatch complaints and enquiries
are received in the latter fashion).
Ofgem requested information, which we provided,
on the 239 complaints that related to actual disconnections, of
which only 38 related to the period between October 2003 and March
2004. The principal reason there were so few in this period was
that British Gas, by far the most prolific disconnector previously,
had imposed a moratorium on disconnections following the deaths
of Mr and Mrs Bates. As explained in the letter of 9 September
to the Chairman of the Committee from Allan Asher,[17]
Ofgem then chose to analyse these complaints within a very narrow
definition of disconnection in error. This did not include those
complaints in which the consumer had disputed responsibility for
an account only to be disconnected regardless, and had the effect
of further reducing the number to 18.
Therefore, as outlined above, of the total of
239 complaints relating to actual disconnections, 38 were received
between October 2003 and March 2004. Of the remaining 201, 133
related to disconnections which took place between April and October
2003, and 68 to disconnections which took place prior to April
2003. The reason this latter figure was included in our evidence
was due to a time delay between the disconnection itself and the
consumer contacting energywatch.
Finally, the proportional analysis in section
2.2 of our memorandum to the committee is based on the 239 incidences
of disconnection in error.
I hope that this letter has been sufficient
to provide the necessary additional information in the areas in
which the Committee was seeking clarification. Of course, if I
can provide any further information please do not hesitate to
contact me.
20 October 2004
(a) the Authority; or
(b) the holder of a gas licence or an electricity
licence,
to supply to it, in such form as it may reasonably
specify, such information specified or described in the direction
as it may require for the purpose of exercising its functions.
(2) A person to whom a direction under this section
is given shall comply with it as soon as is reasonably practicable."
16 "24.-(1) The Council may direct- Back
17
Not printed. Back
| 
