What is the issue with land registration?

When we think about land registration, we tend to do so in terms of title. The passage of Sir Robert Richard Torrens' Real Property Act in South Australia in 1858 introduced title by registration which has been widely adopted around the world, at least in common law-based jurisdictions.

Torrens essentially replaced registering by deeds with title registration. This approach has several well-documented benefits, such as dealing with the problems of poor indexing associated with 19th century paper-based practices.

However, it also has some limitations, which are less well known. It is worth stating (spoiler alert) that these do not relate to geospatial surveying.

The issue is not with certainty per se, rather with indefeasibility; that is, immunity from competing claims. The problems are whether indefeasibility is immediate or deferred, and how to correct errors.

The 1858 Act, commonly referred to as Torrens, has three defining principles.

These three principles provide the foundation for the single most important quality, indefeasible title, which provides an owner with immunity from attack by adverse claims.

In a pure Torrens title implementation, indefeasibility is immediate, and there tends to be limited scope for a party to regain rights that it has been deprived of by fraud or error. Exceptions to indefeasibility are usually set out in statute, and vary by jurisdiction. But if we were designing a title registration system now, would we do it the same way?

The Torrens approach gained traction because it simplified the transfer process and gave certainty about outcome. However, this simplification comes with costs, the most significant being an erosion of the high degree of ownership security afforded by deeds. As summarised by Kenneth Reid of the University of Edinburgh's School of Law in 2016, 'to make life easier for acquirers is also to make titles to land less secure'.

This is not to imply that title by registration is wrong; but we would like the need for change at least to be debated. There should also be discussion over how to overcome some constraints that title by registration places on registrars, and whether these are still necessary in the current social, economic and technological climate.

Originally, registers of land merely recorded deeds, the legal facts of a transaction. Each deed is evidence that a particular transaction took place. For a land register, a registrable deed describes changes to party, rights or land relationships. But in itself it does not prove title; it is simply a record of an isolated transaction, describing a rights change.

A person buying land on a deeds register can only verify the seller's title by examining a sequence of previous deeds, known as the title deeds or chain of title. The legal facts are reinterpreted for every transaction. This means that the derived title reflects the current chain of evidence – accounting for rectified and voided deeds – and are interpreted according to current legislation, so that the derived state reflects the most current legislative mandate. This can be time-consuming and costly.

Title by registration went further by directly registering the legal consequences. In this approach, the title is the primary legal object. This means there is no need for previous supporting deeds, because the title could contain all the relevant legal information on the register.

In fact this is not the case, since the register does not represent overriding – that is, off-register – interests, which are an increasingly common tool for jurisdictions to impose policy restrictions on landowners, such as building, pollution and environmental control which are managed by government agencies outside the Land Register. Title by registration means that any changes to parties, rights or land described in a deed are registered. Interpretation of the transaction resides with the registrar, and is done once at the point of registration.

In title by registration systems, the consequences, or title, become the legal reality rather than the underlying facts, or deeds. The title is always legally correct, even if we know that the underlying deeds would make it incorrect. This is known as a bijural inaccuracy, where the supporting registration law – defining title – overrides general property law. But as Dutch scholars Jaap Zevenbergen and Hendrik Ploeger point out, this 'seems to be an inversion of the original intent of the mirror principle'.

In summary, a register of deeds is a record of facts about change, while a register of titles expresses the consequence of this change. Deeds show the working but leave the reader to interpret the results, which is influenced by prevailing social and legal conditions. In contrast, a register of title shows the results without the working.

A register of deeds system affords the most protection to the true owner because it is based exclusively on the law of property. The problem under Torrens is that the certainty may be granted irrespective of the underlying legal facts of the deed. A title by registration system reduces the security of the true owner, given that errors – including fraud – can become indelibly registered because common law is overridden by registration.

In other words, if a mistake is made on the register, property law can make it impossible for the rightful owner to reclaim their land. Under the curtain principle, it is not possible to reinterpret the deeds – that is the chain of title – while under the insurance principle, monetary compensation is often the only remedy available.

Although many see the benefits of title by registration as self-evident, it is not universally embraced. For example, papers from a conference on land registration and title security in the digital age offer a critique of the system. There are thus many jurisdictions that provide security of tenure exclusively using registers of deeds – for example, France, South Africa and the Netherlands.

In the main, there should be no difference between how rights are resolved between the two types of register. The issue with a system of title by registration is what happens when an inaccuracy occurs.

Some of these issues have played out in Scotland. After nearly 400 years with a register of deeds, the nation implemented a register of title. Just 33 years later, however, it has now introduced a hybrid between a titling and deeds system. Clearly title by registration alone has not been ideal there.

Modern databases offer significant advantages in terms of indexing. Spatial databases and associated spatial indexing (the ability to rapidly identify rights relationships between different features through their spatial geometry) are particularly powerful.

Three key areas need considering:

Digital land registers offer new opportunities, though they face different challenges. The technology is mature and provides clear benefits. Furthermore, if the technology provides a more effective and efficient register, what impact will this have on modern consumer, policy and societal expectations?

There are strong arguments for change. Lawmakers recognise that land registers are facing new challenges, and are drafting legislation that repositions their social and political role. Registers are, after all, a reflection of a social need – and this does change, albeit slowly.

Scotland's hybrid approach to registration combines the benefits of deeds and title approaches in a single framework. A future article will explain how we can build a fit-for-purpose digital register that combines both legal and technological innovations that will lead to new ways of thinking about land registration.

Duncan Moss FRICS is a principal consultant at Ordnance Survey
Contact Duncan: Email

Dr Anthony Beck is land administration lead at Ordnance Survey
Contact Anthony: Email

Related competencies include: Access and rights to land, Land administration and cadastre, Legal/regulatory requirements