Could I get that in writing?

June 18, 2018

Terms seeking to restrict the manner in which contracts
may be varied are extremely common. In my experience they are almost inevitable
in major IT projects – inserted in the hope that they will promote certainty as
to the terms on which the parties are dealing at any given time, and their
respective rights and obligations. Many organisations insert such clauses into
their standard terms for similar reasons. Often, however, they run squarely
into the realities of on-going projects. Timescales are amended, requirements
are tweaked, and often it is assumed that commercial issues can be resolved
later. However, despite the ubiquity of anti-oral variation clauses, there has
been long-standing doubt as to how effective (if at all) they were. Last year
the Court of Appeal appeared finally to resolve it, by confirming what had
generally been the indications in the decided cases: such clauses were generally
ineffective to prevent non-compliant changes to contracts (see the
article by my colleagues at 4 Pump Court
in the February/March 2017 edition
of Computers & Law).

The Supreme Court has now reversed that decision, in a
ruling with very significant implications for those involved in IT projects.
The case was Rock Advertising
Limited v MWB Business Exchange Centres [2018]

Disputed clause

The clause facing the Court provided that:

“All variations to this Licence must be agreed, set
out in writing and signed on behalf of both parties before they take

The dispute involved a licence to occupy premises. The
licensee got into fee arrears, and an employee of the landlord orally agreed to
a staggered series of repayments. However, the landlord subsequently disregarded
this agreed rescheduling and evicted the licensee. 


The Supreme Court concluded
that the landlord was entitled to do so: the anti-variation clause meant that
the renegotiated payment schedule was not binding.

The Supreme Court identified that the traditional
objection to anti-oral variation clauses was a conceptual one. English law,
save in exceptional circumstances, does not impose formal requirements on the
making of contracts, nor on their amendment. Thus, if parties wished to agree a
variation orally, they would be free to do so, and would implicitly agree to
waive the anti-variation provision as they did.

The Supreme Court rejected this view for several reasons.
Firstly, the Court recognised that anti-oral variation clauses serve a
legitimate and understandable commercial purpose. Businesses continued to
include them in contracts despite their uncertain legal status, and the law
ought not to determine that such agreements should have no effect in the face
of the apparent wishes of the parties. Secondly, the majority found that there
was no contradiction between freedom of contract and the freedom to restrict
future changes to an agreement: respecting their determination as to how
changes were to be controlled acknowledged the parties’ autonomy rather than
undermining it (Lord Briggs dissented on this point, though not on the ultimate

The Supreme Court also observed that, generally, parties
agreeing oral variations are not consciously agreeing to dispense with anti-oral
variation provisions. They are generally forgetting such provisions exist. This
is certainly my experience.

The argument against such a change is the injustice which
might well result from allowing a party to go back on an agreement as to how a
contract ought to be performed. The obvious answer is that estoppel will step
into the breach to prevent injustice. As many reading this will know, estoppel
is the doctrine by which courts will prevent a party from going back on
representations it has made to another party in circumstances where it would be
unjust to permit it to do so (generally because the other party has relied on
the representation to its detriment). It is easy to imagine many situations in
which this might operate to prevent a party which has permitted, or even
encouraged, a counterparty to perform differently from the way required by the
contract from subsequently seeking to accuse such a party of breach of

However, the Supreme Court gave a clear warning in this
regard. It stated that estoppel could not be allowed to operate so broadly that
anti-oral variation clauses were deprived of any real effect. As such, the bar
to establish an estoppel will be high. The Supreme Court said that a party
seeking to establish an estoppel will need to identify “words or conduct
unequivocally representing that the variation was valid notwithstanding its
informality”. That is clearly a high bar, and it was made expressly clear
that simply entering an agreement notwithstanding the anti-oral variation
clause would not suffice. Nor, the Supreme Court made clear, did having paid lesser
sums than those owed to a new timetable amount to a necessary detriment to
found an estoppel or make revisiting the agreement unconscionable.


The approach of the Supreme Court will doubtless be
welcome to those tasked with drafting and policing performance of IT contracts.
Anti-oral variation clauses have an obvious commercial rationale yet, until
now, have been of limited and doubtful effect. Parties can now anticipate being
held much more closely to what they have agreed at the outset, and lawyers can
sleep somewhat better at night knowing that the risk of employees varying
contracts without proper authorisation to do so (or informing the lawyers of
what is happening) has diminished. In theory, arguments about what contractual
obligations are or were mid-way through a project (with contractual dates
missed and the parties having adopted, worked to and revised a number of plans
in the interim) should be easier to predict and simpler to resolve.

However, there seems to be a real risk of injustice and a
need for heightened vigilance if those involved in projects are not to be
seriously caught out. The first message is one familiar in these pages: parties
who put the contract in a drawer at the start of a project and return to it
only when deep in dispute run a serious risk: it will be important to be
conscious of what the existing contractual commitments are and the requirements
to change them. If changes are genuinely desired by both sides, formal
processes will need to be complied with. Perhaps uncomfortably, suppliers will
need for their own protection to insist on formal changes (or an unambiguous
acknowledgement that such a requirement is waived) and steel themselves for the
difficult conversations which may accompany them. Where previously a little
ambiguity and fudge may have eased difficult relationships along, that is no
longer likely to be an acceptable approach given the risks involved.

An interesting aspect for the future will be how the courts
approach claims of estoppel in light of the Supreme Court’s comments. It seems
to invite considerable injustice if both parties can agree (informally) to a
change and act on it only for either party to be free to insist that the change
was of no effect and, on the contrary, opens the other to a claim for damages.
The Supreme Court strongly implies that this will be the outcome unless there
is express acknowledgement of the anti-variation provisions and the fact that
they are being waived. However, it seems probable that first-instance tribunals
will seek to avoid injustice as far as possible, through a generous approach to
finding compliance with formal requirements or refining the requirements of an
estoppel in such circumstances. Given the ubiquity of anti-variation clauses,
we are unlikely to need to wait long to see these questions further explored in
the courts.


Richard Osborne is a barrister specialising in information technology disputes practicing at 4 Pump Court. He is ranked as a leading junior in IT and telecoms by both Chambers & Partners and the Legal 500.