How to evaluate the reliability of economic data?

Readers Question: how to evaluate the reliability of economic data?

This is a good question. In brief, some issues to consider

Is there a political corner? – Political inclination of the media / think tank

  • In what context are the data presented? How to interpret the data. If we consider nominal data, real data or a percentage of GDP, each statistic will give a different impression.
  • Is the data used to support a particular opinion? – Does the publisher search for data that supports a particular opinion? For example, if you want to be critical of the 1980s, you can find data that shows a rapid decline in production work. If I wanted to give a more optimistic impression, I could find data on the time lost due to strikes
  • Who publishes the data? Do the statistical authorities have independence and good results? For example, countries like Argentina and China have been accused of bringing data like inflation for political reasons below. UK ONS has a reputation for independence.
  • How do you compare data with other statistics? An example, in the United States, there is criticism of US inflation data (from organizations like “Shadow Stats” that claim that the real inflation rate is much higher because the American statistical agency ignores the changes in the product quality. However, the billions of prices project “suggests an inflation rate very similar to that of the official agency. See” Billions of prices cannot be wrong “by Tim Harford in the FT
  • How is the data collected? Does it come from a small survey, which may have sampling errors or is it a national survey?
  • How selective is the data? The choice of particular years can give a particular impression. For example, data from the 1950s and 1960s suggest that the relationship of the Phillips curve works. But, if you choose the 70s, the relationship is interrupted.
  • How to use the data? Does an increase in the budget deficit mean that the government has failed or is it a correct response to an economic recession?
  • Who should attribute credit / guilt to the data? US presidents are often judged on economic data, but to what extent are they really responsible?


What is included in the data? Recently the UK government made the claim that –

"The United Kingdom is the third country that spends more on education in the OECD, and now there are 1.9 million more children in" good "and" exceptional "schools compared to 2010". (1)

This is a statistic used by the government to reject requests for “cuts” in spending. However, critics argue that the data is misleading because this statistic includes private spending on tuition and universities. Therefore, at best, the statistics are a misleading guide to public spending in high school.

According to the School School, it has re-recorded the data on the OECD website, so that only public spending on education has been counted as a percentage of GDP in all countries. In this metric, the United Kingdom fell from third place to 15 for education expenses as a percentage of GDP, out of the 36 member countries.

Spending per capita

Another problem is that the government aims to spend on higher education and more students in a good school. However, the UK is also experiencing population growth. Nominal increases in expenditure are misleading if the number of students involved is excluded. A more useful statistic is to measure spending per student, as this statistic is better translated into educational resources per student.

Likewise, spending on medical care is subject to similar uncertainties.

Actual expenditure on medical care has increased, but as a percentage of GDP it has decreased.

Worst level of UK debt on record

This graph shows the total debt level of the UK government. It shows a rapid increase since 2002. It is true that the latest figures show a record level of total UK debt. However, at best, it is a misleading figure. Total debt levels need context. What percentage of GDP? What percentage of tax revenue? If the GDP increases at the same rate as the total debt, the debt is manageable.

If we look at debt as a % of GDP, then the increase is smaller and even tails off in 2016.

This may seem like a record level of debt as % of GDP, but it is also useful to put in a historical context.

In this context, the increase in debt as a percentage of GDP seems less dramatic.

Another problem is how to use the data. A newspaper may report rising debt levels by assuming that this is a “bad thing”, but from a Keynesian point of view, an increase in public debt during a recession would be considered a necessary or even beneficial response to the recession.

Isolated data
A quarter of economic growth is always an unreliable guide to judging economic performance.

Leave a Reply

Your email address will not be published. Required fields are marked *