Digitizing Land Registry – are we doing enough?

I am not sure if I am the only person who has noticed that the first task each new Lands minister does upon appointment is to shut down the Lands registry in order to carry out some digitization exercise.

I would start off with a disclaimer by saying that I am not privy to details of the current digitization exercise, but am pretty sure the previous digitization exercises have not successfully stopped fraudulent transactions that continue to plague our land registries.

Those who keep track of global financial scandals will remember the Enron Scandal where auditors helped a multinational cook their financial books in order to prop up its stock valuation.

Enron financial books were as digitized as any blue chip company would want them to be. But through ‘creative accounting’ or collusion between key actors, they ensured that key financial data regarding their huge debt portfolio was hidden from the public.

Basically the company was insolvent, but the auditors and accountants propped it up on the stock markets as being very healthy financially. This point illustrates the fact that whereas digitization improves efficiencies, it is never sufficient to prevent fraudulent activities.

This is because the current models for digitization or computerization are based on what is known in technical jargon as ‘Client-Server’ systems. In Client-Server systems, all participants in the digital transactions must place trust in whoever is controlling the main computer or server.

If Akinyi sells land to Biketi, the digitized system would increase efficiency by allow her to complete the transaction online. However, nothing other than ‘trust’ will stop the personnel who manage the servers at the Lands registry from fraudulently modifying that record to show that Akinyi ALSO sold the same land to Chebet.

Effectively, Biketi and Chebet will be holding claim to the same title, but they wont know about it.

Furthermore, the different banks that each owner will approach for a construction loan by using their title as collateral will also not know about the fraudulent situation occasioned by the ‘trusted’ entity at the ministry of Lands.

This implies that digitizing the lands registry requires more than putting land titles and their owners online. It must go deeper to address the fundamental question of where we should place the trust to authenticate and validate transactions.

Fortunately, there are new technologies that are designed to democratize the trust we have traditionally placed one centralized agency by re-distributing it across several stakeholders.

Essentially what this means is that when Akinyi sells her land to Biketi through the digitized system, that transaction will need to go through a democratic process of verification and validation as prescribed by the group of stakeholders.

Each of the stakeholders will therefore independently verify and validate the transaction by running a piece of code that checks if Akinyi is indeed the land owner, and that she is the one who has authorized the sale to a particular buyer.

The transaction is only entered into the lands registry once all or majority of the stakeholders are in agreement. This raises the integrity bar in that it would require a majority of stakeholders to fraudulently modify the transaction - as opposed to current situation where just one of them can execute the fraud.

There are several key stakeholders that are impacted by land transactions and are therefore good candidates for participating in such an automated, consensus type of validation.

Beyond the Lands ministry, other stakeholders would include the Lands commission, the Law Society of Kenya, the Banks, the Kenya Revenue Authority, and the County Governments.

Each of them could be tasked to install and ran an independent validation node that would participate in the consensus protocol in order to increase the integrity of the land transactions.

Without decentralizing and distributing the verification and validation process in the lands ministry, we face the risk of digitizing fraud rather than eliminating it.

==ends===