2013-08-13 04:57:38
Aug 8th 2013, 11:40 by C.W. | LONDON
HOW much we spend depends on how old we are. Generally, people increase their day-to-day expenditure until middle age, but spending drops thereafter. Economists call this the hump in life-cycle spending. But it is difficult to determine exactly what drives the fall in spending as people age. Most people think that declining income leads to cutting back on non-essential purchases. This makes intuitive sense. If you have less purchasing power, it is easier to cut back on holidays than it is on food.
A new paper* challenges this consensus. The authors, Mark Aguiar from Princeton and Erik Hurst from Chicago, argue that our understanding of life-cycle expenditure has missed quite a bit of detail. The focus of their paper is on non-durable items goods that wear out quickly, such as food, clothing and drink. The authors start their analysis on familiar ground. They show that expenditures peak in middle age, at a level roughly 25% higher than expenditures at 25 or 65: the classic hump shape.
But their data, primarily drawn from the Consumer Expenditure Survey from 1980 to 2003, provide an alternative explanation for why expenditure falls as people enter old age. To start with, they reveal that spending on non-essential items does not drop. In fact, it increases. But three categories do see declines: food, transportation and personal care (which includes clothing).
To explain these trends, Messrs Aguiar and Hurst argue we cannot just look at fluctuations in financial firepower. Instead, lifestyle changes are more important. For example, beyond the age of 60, when people are more likely to be retired, expenditure on transport to work declines. Non-work travel time actually increases over the second half of the life cycle; people have more time to go to museums and see friends. But, overall, the effect of retirement is so strong that total travel time, and spending on travel, drops. The authors make a similar argument in relation to money spent on clothing. Quite simply, people out of work need fewer or less expensive clothes.
The authors also argue that, as people age, their relationship with time changes. Economists are fond of the phrase opportunity cost , which refers to what you forego in order to do something else.
For people in work, the opportunity cost of time is high. An hour spent preparing a meal at home could be an hour s foregone earnings; it might make economic sense to go to the local takeaway instead. But when retired, people might not think in this way. Rather, they may take pleasure in spending time making a quality meal, rather than buying it. And the authors find that the decline in expenditure on food consumed away from home after middle age is driven by people visiting fast food establishments or cafeterias less frequently. In addition, after middle age, individuals spend more time carefully preparing meals and shopping for food, leading to lower expenditure on food at home.
But the authors show that elderly people do not spend less at table-service restaurants. These restaurants do not serve a time-saving function to anybody; rather, they are pleasant places to go to. So, even though retired people have more time, they are as inclined to sit down for a meal at a nice restaurant as they ever were.
The paper shows that changing time richness, rather than income richness, is what drives changing spending patterns in the elderly. Why are these findings important? They could revolutionise the way that we measure poverty. Normally, expenditure is the measure used to assess whether someone is in poverty or not. And by this metric, older people are more likely to be in poverty. The elderly do spend less. But once we include time into the equation, grey poverty levels might be lower. Elderly people have more time, and this can compensate for lower income.
The problem for researchers is that government data collection practices lag behind advances in economic thinking. Mr Hurst grumbles that most methods for wellbeing such as the British Household Survey typically do not include many questions about time use. But it would be relatively easy to include them. And this would help us to understand better exactly what happens to people as they age.