VARC Practice Set 2

On a summer afternoon, the surface of an asphalt car park can be thirty degrees hotter than the air above it, and a city centre can be several degrees warmer than the countryside around it. This urban heat island is caused less by the heat cities generate than by the heat they store: concrete and asphalt absorb sunlight by day and release it slowly through the night, so that cities cool less after dark. For residents, the nights matter most. A hot day is uncomfortable; a run of hot nights, with no chance for the body to recover, is when heat becomes deadly, especially for the elderly.

Trees are the most widely recommended remedy. A mature tree shades the ground beneath it and cools the air through transpiration, releasing water vapour from its leaves much as sweat cools skin. Streets lined with trees can be markedly cooler than bare streets nearby. Many cities have announced ambitious planting targets, often expressed as a number of trees or a percentage of canopy cover.

The targets can mislead. A sapling provides almost no shade, and in many cities a large share of newly planted trees die within their first few years, from drought, vandalism or simple neglect. A programme that counts trees put into the ground rather than trees that survive to maturity may report success while delivering little. Moreover, canopy is unevenly distributed. In many cities, wealthier neighbourhoods have far more of it than poorer ones, partly because of history—older suburbs were laid out with tree-lined avenues—and partly because affluent residents are better able to lobby for street trees and to care for them.

None of this argues against trees. It argues for treating them as infrastructure rather than decoration: budgeting for watering and care over decades, measuring outcomes in shade delivered rather than saplings planted, and directing effort first to the neighbourhoods where heat does the most harm. Trees take a generation to pay their dividends. A city that wants cooler nights in thirty years has to start caring for its trees now, and keep caring for them long after the planting ceremony is forgotten.

A cafeteria that places fruit at eye level and desserts at the back can change what people eat without banning anything or raising any prices. Behavioural economists call such interventions nudges: alterations to the way choices are presented that steer people in a predictable direction while leaving all options open. Since the idea was popularised, governments have set up units to design nudges for everything from tax compliance to organ donation.

Supporters argue that nudges are unavoidable. Some arrangement of the cafeteria has to exist, and every arrangement influences choice; the only question is whether the influence is deliberate and benign or accidental and possibly harmful. If a pension scheme must have some default contribution rate, it might as well be one that serves most employees well. On this view, objecting to nudges is like objecting to the existence of menus.

Critics reply that the inevitability of influence does not settle the question of who should exercise it, or how. A nudge works best when people do not notice it; a default is effective precisely because most people never examine it. This makes nudges hard to contest in the way that laws can be contested. A tax can be debated in parliament; the placement of a checkbox on a form is decided by officials and seldom discussed. Critics also worry that nudges treat citizens as predictably irrational subjects to be managed rather than as reasoners to be persuaded.

A middle position holds that nudges are legitimate when they would survive publicity—when officials would be willing to announce them openly and when people, told about them, would broadly endorse them. A default that enrols workers in a pension unless they opt out passes this test; most people, asked, approve of it. A design that exploits confusion to sell an unwanted product would not. The test does not eliminate disagreement, but it gives it a place to happen.

In societies without writing, knowledge survives only if people remember it. Genealogies, laws, navigation routes and the properties of plants must be carried in heads and passed on through speech. Outsiders long assumed that such memory must be unreliable, a game of whispers in which details shifted with each telling. Recent work suggests that this assumption underestimates the techniques oral cultures developed to make memory durable.

Many of these techniques attach information to places. An Aboriginal Australian songline, for instance, links a sequence of songs to features of the landscape, so that walking the route, or imagining walking it, cues each verse in turn. The classical memory palace used by Greek and Roman orators worked on the same principle: a speaker would imagine placing each point of an argument in a room of a familiar building and then mentally walk through the building as he spoke. Spatial memory, which evolved to help animals find food and shelter, turns out to be an excellent scaffold for abstract information.

Other techniques rely on form. Rhythm, rhyme and fixed formulas constrain what can be said, so that an error disrupts the pattern and is more easily noticed. Group performance adds a social check: when a song is sung by many voices, an individual's slip is corrected by the rest. Some communities assign responsibility for particular bodies of knowledge to particular families, who are expected to reproduce them exactly.

Such evidence does not prove that oral traditions never change. They do, sometimes deliberately, as communities revise stories to fit new circumstances. But it suggests that change is not the same as decay. A written text is fixed but can be lost or misread; an oral tradition is fluid but is constantly rehearsed and checked by those who depend on it. Each medium has its own vulnerabilities, and the prejudice that writing is simply superior says more about literate societies than about memory.

Much of the software that runs the internet was written by people who were not paid to write it. Web servers, programming languages, encryption libraries and operating systems are maintained by volunteers or by employees whose companies allow them to contribute. To an economist of an earlier generation this would have seemed impossible: why would anyone produce a valuable good and then give it away?

Several answers have been offered. Programmers contribute to build reputations that later help them find jobs; companies contribute because they rely on the software and benefit if it improves, while sharing the cost of maintaining it with rivals. Some contributors are motivated by the pleasure of solving problems or by a belief that knowledge should be freely shared. These motives together have sustained projects of astonishing scale.

Yet the model has a weakness that has become more apparent as software has become more essential. Contributions flow towards exciting new features and away from tedious upkeep. A widely used library may depend on one or two maintainers working in their spare time, and when a serious flaw is discovered, the world learns how much rests on so few people. Companies that use such software freely often contribute nothing back, a classic problem of free riding: each user benefits whether or not it pays, so few choose to pay.

Proposed remedies vary. Some advocate funding foundations that pay the maintainers of critical projects; others want large companies to commit a share of their engineering time to the software they depend on; still others propose that governments treat key open-source components as public infrastructure, like roads. Each remedy raises new questions—about who decides which projects are critical, and whether payment might change the culture of voluntary collaboration that made open source succeed in the first place.