Economics
Using numbers and statistics in reporting
1) Avoid using multiple figures and statistics in quick succession. (A good rule of thumb is one per sentence — whether that’s in a written story or a broadcast script).
DO THIS
A report on U.S. wage growth includes the following sentence: “As of November, wages for part-time and full-time workers were 6.2% higher than in 2021.”
DON’T DO THIS
An article on wage growth includes the following sentence: “Last month, average hourly earnings rose by $0.09 to $32.82. That's a 53% decline in the pace of wage growth from December 2021 when average hourly earnings jumped by $0.19. On an annual basis, average hourly wages are up 4.6%, that's one-tenth of a percentage point below the annual rate at the end of 2021.”
DISCUSSION
Especially when moving quickly or responding to breaking news, it can be easy to present numbers as if they are the story — even if they draw upon many different mathematical concepts (including magnitude and scale, proportions or percentages and comparison). The second example includes three different quantitative indicators, and assumes pre-existing knowledge of the concept of “medians.” This is simply too dense for a reader, viewer or listener to understand. Look for ways to translate the data, too — for example, “The highest-paid workers got a much bigger raise than the typical worker.”
2) Identify where the number comes from
DO THIS
A report on the public school system in Kansas observes that “there are about 1,400 teacher vacancies, according to data from the state department of education.”
DON’T DO THIS
A report on the public school system in Kansas claims that the state is “facing the most severe teacher shortage it’s ever known,” but provides no source to support the claim that 4% of teaching jobs (1,400 positions in all) are unfilled.
DISCUSSION
Letting audiences know where cited numbers and statistics come from will help them think critically about economic data. Even language as simple as “study claims” or “according to the…” is helpful here, as it will help audiences identify the source of a given data point, and communicate the point that numbers are not inherently “facts.”
3) Numbers don’t speak for themselves. Provide a concrete context for understanding its significance.
DO THIS
DON’T DO THIS
“Japan's core consumer prices rose 4% in December from a year earlier, government data showed on Friday.”
DISCUSSION
When personalizing economics or business reporting, the goal should always be to use numbers and statistics to flesh out a story, not to assume that those numbers and statistics are the story. One of the best ways to help the public understand data is through the use of real-world examples — as in the rising cost of “everything from burgers to gas.” The inclusion of this phrase in the first example helps audiences understand the meaning of the consumer inflation data. Changes in the economy have a human impact. By contrast, while the second example informs readers that there is something unique and somewhat unprecedented happening with Japan’s consumer price index (as the phrase “fastest pace since 1981” indicates), this sentence does not explain what that feels like in real-life terms.
4) Always explain the measures of economic indicators you are using and how they are reached. Give readers a sense of what the numbers don’t show, or where they can fall short.
DO THIS
A report citing unemployment rates for a number of Kentucky counties explains “Unemployment statistics are compiled to measure trends rather than actually to count people working. They do not include unemployed Kentuckians who have not looked for employment within the past four weeks. The data should only be compared to the same month in previous years.”
DON’T DO THIS
A report describing why Shelby County, Alabama, has the lowest unemployment rate in the state says “The reason for the county’s low unemployment numbers is their large participation rate, meaning more people are working.”
DISCUSSION
Indicators like the unemployment rate can give us a useful perspective on the economy, but that single perspective does not paint the full picture. Draw attention to what economic concepts overlook or leave out, so audiences can better understand their benefits and limitations. By explaining that the unemployment rate excludes those who have been out of the labor force for more than a month, the first example helps readers understand some of the disadvantages of this metric. By contrast, the second example suggests that the unemployment rate is a function of the labor participation rate — an entirely different kind of measurement.
5) Whenever possible, use human stories to illustrate economic realities. Avoid abstractions like “U.S. consumers,” “the middle class,” or other ambiguous groups of people.
DO THIS
An article called "Inflation is Battering Low-Income Households Most" compares the impact of rising prices on low-income and middle-income households, and explains what these terms mean (for example, low-income households = “the bottom 40% of households by income”).
DON'T DO THIS
An article called “15 Ways the Middle Class Can Afford a Trip to Disney World” provides tips on how Americans can save money while visiting this popular tourist attraction, but says nothing about who makes up the “middle class.”
DISCUSSION
Most Americans consider themselves "middle class," so they are likely to interpret the term through that lens. Affluent people will assume that the "middle class" is wealthier than it is, while lower-income people will assume that the "middle class" struggles more with finances. Being precise about who you mean helps audiences understand how economic forces and factors influence the lives of actual people.
Both examples would benefit from some human stories, too. When personalizing economics or business reporting, the goal should always be to use numbers and statistics to flesh out a story, not to assume that those numbers and statistics are the story.
6) Share the methods and evidence that led researchers to their conclusions.
DO THIS
A report claiming that the likelihood of a U.S. recession is “on the rise” cites research from the St. Louis Federal Reserve Bank indicating that 27 U.S. states are “showing signs of faltering economic activity,” and then explains how “an average of 26 states displayed negative growth during the last six recessions dating back to 1980.”
DON’T DO THIS
A report on the world economy notes that a probability model has predicted a “98.1% chance of a global recession,” but says nothing about how this result was arrived at nor the methodology that informed the model’s creation.
DISCUSSION
Helping audiences understand the basis for a claim can go a long way toward increasing their knowledge of how economies work. As such, when discussing economic research, it’s important to not only share the results of individual studies, but to also talk about how researchers reached their conclusions. In simply relaying results from a forecasting model, the second example fails to do this. By contrast, the first example provides evidence to support the researchers’ argument about the risks of a recession.
7) When discussing economic data, acknowledge how difficult it is to be precise, and emphasize inconsistencies, uncertainties, and unknowns. Don’t imply that things can be known with absolute certainty.
DO THIS
A story notes that 1,857 people in Modesto, California, were officially counted as “homeless” in 2022, and then goes on to explain why these figures are both “imprecise” and “severely flawed.”
DON’T DO THIS
A report on homelessness in Corpus Christi cites specific numbers tallied in different years, but never mentions their accuracy or the difficulty in counting them.
DISCUSSION
Counting things is often very difficult – in economics as in other areas of life. In many cases, it is impossible to arrive at an exact figure, which is why it’s important to do two things: 1) let audiences know that most numbers and statistics are estimates; 2) show them how incomplete and partial many of our data-gathering methods are, and what they miss. The first example does this quite effectively, while the second implies that the data shared on homelessness is totally accurate.
8) Look at multiple indicators. Don’t reduce the economy to a single data point.
DO THIS
A report describing how the U.S. “added a sizzling 517,000 new jobs last month” notes that while unemployment dipped to 3.4% (“the lowest level since 1969”), these trends have also “contributed to high inflation.”
DON’T DO THIS
An article discussing how the U.S. “added 517,00 new jobs last month” notes that this “push[ed] the unemployment rate down to 3.4%,” and then goes on to say that this job growth “def[ies] fears that the economy is heading for a downturn.”
An article discussing recent stock market trends begins with a headline reading “The Economy is Flashing Red: Be Very Careful with Your Stocks.”
DISCUSSION
It is overly simplistic to look at economies solely through the lens of unemployment statistics or any other individual indicator. As the second example indicates, just because one part of the economy (in this case, the labor market) is looking strong doesn’t mean that the entire economy is strong. This is the second example’s unfortunate message, which mistakenly informs readers that job growth indicates an economic upswing. To avoid these kinds of reductionist views, place different kinds of indicators in conversation with each other. Help audiences understand the different kinds of measurements that are used to assess the health of an economy. That includes the stock market. The stock market is not the only measure of the economy, and equating the economy as a whole with the stock market has the potential to exclude a lot of people: Just over half of Americans have any investment in stock, including mutual funds or retirement accounts. Just as the economy is much more than the stock market, the stock market itself is much more than the Dow Jones Industrial Average. To report with a more holistic, comprehensive view of the stock market, look beyond the Dow and consider how other major indices are performing.
9) When sharing economic data, don’t assume that things are the same everywhere. Look for ways to highlight differences between different groups, between different regions, and between different points in time.
DO THIS
An article on food prices notes that the impacts of inflation depend “on your location and income level,” and draws attention to a study that revealed “stark regional differences” in food spending.
DON'T DO THIS
Another article on food costs explores increases in the price of various goods, but focuses only on national data provided by the Bureau of Labor Statistics.
DISCUSSION
It is simply not possible to count or check all prices everywhere, so they will always be an estimate. There is a lot of regional fluctuation in prices, as well as seasonal changes (turkey cost on average $1.81/lb in October 2022, but many grocery stores cut prices before Thanksgiving). National-level reporting may not fully represent the reality that individual consumers face at the store.
10) Don't attribute the state of the economy to a single politician
DO THIS
A headline reads "The US economy’s growth was stronger than expected in the third quarter."
DON'T DO THIS
A headline reads "Biden's Economy Has the Best Growth Record Since Clinton."
DISCUSSION
Politicization can make it harder for audiences to focus on the topic at hand. And it's inaccurate, to boot: presidents don't have that much control over the economy, and their decisions can take years to have impact. Citing years (ie the 1990s) instead of Clinton would be more useful, unless you’re talking about specific policy.
11) Look for ways to independently scrutinize official statements. Don’t just repeat what governmental agencies or elected officials say
DO THIS
A report on the number of jobs created through various economic projects compares official projections to historical realities, and finds that few of these projects “lived up to their job creation goals.”
DON’T DO THIS
A report about a new manufacturing site begins with a headline stating that “Qcell will create 2,500 new jobs in Northwest Georgia.”
DISCUSSION
Government agencies and officials commonly make announcements and pronouncements about the economy. It’s important to do more than just passively repeat these. By using an official study to evaluate the accuracy of job creation projections, the first example shows how economic data can be used to support both statistical literacy and critical thinking.
- Previous
- Constructing headlines