Real-time data and new economic frameworks are transforming the decision-making processes across domains. This year's Budget and Economic Survey used terms like Barbell Strategy, Agile Response, and Bayesian updating of information to tell how the government has improved its decision-making in response to Covid-19.

We decode these terms to understand how new methods are transforming governance.

Barbell strategy

First the Barbell Strategy. A barbell is a weight lifting equipment consisting of a long bar with equal weights attached at each end. If one side is heavier, the weightlifter may lose balance and break his spine.

In his book, the Black Swan, Nassim Nicholas Taleb introduced Barbell Strategy. It is “a method that consists of taking both a defensive attitude and an excessively aggressive one at the same time, by protecting assets from all sources of uncertainty while allocating a small portion for high-risk strategies.”

Black Swans are highly improbable events that could be both negative and positive. 9/11, the financial market meltdown in 2008, and Covid-19 are a few examples.

On one end of the barbell, one is highly risk-averse and therefore neutral to negative Black Swan events. On the other end, one dabbles in risky projects, where one can win big if the Positive Black Swan event strikes. Taleb argues, balancing both ends of the barbell this way, one hopes to become "antifragile."

In the economic policy context, the Barbel strategy balances risk and uncertainty. In the last two years, amid Covid and global supply chain disruptions, the government has used 'Barbell Strategy' by providing (I) safety-nets to cushion the impact on vulnerable sections of society/business and simultaneously (II) flexible policy response using real-time information to implement the most forward-looking option available.

Barbell Strategy is extensively used in financial markets and businesses. Jeff Bezos’s anchoring his business around Amazon and also taking bets in space is an example.

A retail investor secures 80 per cent of his capital by investing in FDs and invests the remaining in speculative stocks or crypto-currencies. If Bitcoin goes to zero, he loses just 20 per cent of the money. But a few lucky ones saw the Bitcoin returns up by hundred times. The idea is to focus on extreme ends and avoid medium-risk strategies.

Agile approach

The second concept is an 'Agile Approach' to decision making. Barbel strategy depends on high-quality real-time data. Data on GST collections, digital payments, satellite photographs, cargo movements, toll collections are now real-time. Policymakers use these to tweak the response. They do it by periodically assessing the outcomes of earlier decisions and data on evolving situations. This data-led tweaking of response is called the Agile framework.

The government used Agile Approach last year by first making food available to the poor and providing liquidity/credit support for MSMEs. After securing the most vulnerable sections/businesses, the government worked on targeted development packages for infrastructure, production linked incentives schemes, etc. One can tweak the response further as new data/evidence pours in.

This is a significant improvement over earlier days when limited data and data lag of a few months and years were major policy handicaps. Our five-year plans and most of the world worked on the ‘Waterfall approach’. The approach used available data for planning of outcomes for next few years, and ensured meticulous implementation. There was a limited scope of mid-way course corrections. Economist Friedrich Hayek termed ‘Waterfall approach’ approach "The Pretence of Knowledge”. But there were no better options in the past due to limited data.

Third concept is Bayesian updating of information. Barbel strategy uses the agile framework for revising policy responses in the light of new data and evidence. The Bayesian approach assesses the strength of new evidence to upgrade probability estimates to adjust policy response.

Let's understand the Bayesian approach through an example: There is a prize of ₹1,000 for correctly guessing if it will rain tomorrow. Publically available weather information predicts a 50 per cent chance for rain. Based on this, the general public would be willing to pay ₹500 for a participation ticket.

But Ria, a weather scientist, knows better. She is sure of 75 per cent chances for rains. She would be willing to pay higher for the ticket, up to ₹750. She uses her expertise to improve her probability of winning. But not all such events are impactful from policy angle.

The Bayesian model provides a step-by-step approach to allocating weightage to different evidence. By doing so, the Bayesian approach reduces the effect of cognitive biases. The Bayesian model has given birth to a new class of super forecasters, far better than regular forecasters. Britain used the method to crack the Enigma code during World War II.

New frameworks are transforming businesses and governance in India and worldwide.

AI is a step ahead where the software automatically learns from new information and adjusts response without human intervention. So there is some catching up to do.

Srivastava is from Indian Trade Service; Kamar and Fatima are Research Associates at Department of Commerce, Government of India. Views expressed are personal