Lagging Indicators

Lagging indicators are economic statistics that tend to change direction only after the economy as a whole has turned. They do not predict, they confirm: when a lagging series finally rises or falls, it shows that a shift in the business cycle was real rather than a passing blip. Common examples are the unemployment rate, the average length of unemployment spells, business loan balances, inventories relative to sales, unit labour costs and services inflation.

They are the third of the three timing groups economists use. Leading indicators such as building permits and new orders turn first; coincident indicators such as payroll employment and industrial production move with the economy; lagging indicators follow. The labels describe typical behaviour over past cycles, not fixed laws, and a series can lead in one downturn and lag in the next.

Picture a thermometer stuck in a thick stone wall. When the weather turns cold, the air feels it first, then the outside of the wall, and only hours later does the reading inside the wall start to drop. By the time it does, nobody doubts that winter has arrived.

Lagging indicators behave like that thermometer. Firms usually try to keep staff during the first months of a slowdown, so job losses build up later. Banks keep lending on credit lines agreed months earlier. Prices for services are often set by contracts and wage deals that take time to renegotiate. These delays make the data slow, but they also make them dependable when they finally move.

The Conference Board, which publishes the US leading and coincident indexes, also compiles a Lagging Economic Index from seven series:

• The average duration of unemployment, which keeps rising well into a recovery as people who lost jobs earlier stay out of work.
• The ratio of inventories to sales in manufacturing and trade, which climbs after sales have already fallen.
• The change in labour cost per unit of output in manufacturing, which reflects past movements in wages and productivity.
• The average prime rate charged by banks, which follows central bank decisions on interest rates.
• Commercial and industrial loans outstanding, since companies draw on existing facilities when cash flow weakens.
• The ratio of consumer instalment credit to personal income.
• The change in the consumer price index for services.

Outside that index, analysts often treat the headline unemployment rate, corporate profits, bankruptcies and loan defaults as lagging series.

Three forces explain most of the delay. First, firms hoard labour: hiring and firing are costly, so employers cut hours and overtime before jobs, and rehire slowly after a downturn. Second, contracts and sticky prices mean that wages, rents and many service prices are reset only once or twice a year. Third, balance sheets adjust slowly. A rise in the federal funds rate passes into the stock of loans and mortgages only as old loans are repaid and new ones are written, and loan losses appear only after borrowers have run out of cushions. Long spells without work can also leave lasting damage, an effect discussed under hysteresis.

Lagging indicators matter for policy because central banks are judged on outcomes that lag. Core services inflation and wage growth, both slow-moving, feature heavily in the decisions of the Federal Reserve, the European Central Bank and the Bank of England. The risk is acting too late: by the time unemployment rises sharply, a recession is usually well under way, and policy changes will take many months to work, as explained under policy lags. That is why rules such as the Taylor rule lean on forecasts as well as current readings.

Lagging data also help in dating cycles. The NBER Business Cycle Dating Committee, which sets official US recession dates, waits for broad confirmation across income, employment and spending before declaring a peak or trough, often a year or more after the event. Investors and lenders use lagging measures to check whether a recovery in early signals is feeding into jobs and profits, and credit analysts track default and arrears rates to judge credit risk across a loan book.

The obvious drawback is timing: a confirmation that arrives after the turning point is of little use to someone who needed to act before it. Lagging series can also be distorted by policy. Pandemic-era job retention schemes in the UK and euro area kept measured unemployment low in 2020 even as output collapsed, and debt moratoriums delayed loan defaults. Structural changes, such as the long-term decline in the share of the workforce in manufacturing, can alter how quickly a series responds. Finally, the line between coincident and lagging is not fixed; in some cycles the unemployment rate has moved almost in step with output. For these reasons, analysts read lagging data together with leading, coincident and real-time indicators before drawing conclusions about the macroeconomic picture.

• The Conference Board: US Leading Indicators
• NBER: Business Cycle Dating