Main results of the research

Based on the paper: Media multipliers: Is there a negativity bias in economic news?

1. Newspaper coverage strongly mirrors unemployment dynamics

Figure 1 plots monthly counts of bad news (articles mentioning rising or high unemployment), good news (articles mentioning falling or low unemployment), and their difference - negativity - together with the unemployment rate. Several clear patterns emerge:

• Bad news closely tracks the unemployment rate, with pronounced spikes in the early 1980s recession and the Great Recession.
• Good news move less, rising only modestly in strong labor markets (e.g., late 1990s, post-2015 period).
• The negativity index actually tends to rise months before unemployment does. In other words, the media often signal turning points in the job market before they show up in official numbers  

This happens because journalists talk to businesses, professionals (e.g., academics and policymakers), and workers and often pick up on changes that aren’t yet visible in the data. So the news doesn’t just reflect the economy — it often anticipates it.

Figure 1:

2. Why earlier studies found a “bias” — and why that was misleading

If we use the same statistical test as earlier research — comparing how the media respond to unemployment going up versus going down — we get the same result other studies did: news outlets talk more about rising unemployment than falling unemployment.

However, this doesn’t mean the media prefer bad news.

The key issue is that:

• When unemployment rises, it usually keeps rising for a while.
• When it falls, it often falls only a little at a time.

Because journalists pay attention to where unemployment seems to be heading, increases get more coverage simply because they signal a bigger change in the future. It’s not bias — it’s anticipation.

3. When we control for unemployment’s future path, the bias disappears

To test this properly, we built a measure we call a media multiplier. The idea is simple: instead of comparing the media’s reaction to unemployment this month, we compare their reaction to the whole future path unemployment is expected to take after a shock.

When we do this, the picture changes completely (see Figure 2):

• If we look only at the current month, it appears the media react more strongly to rising unemployment — just like earlier studies found.
• But as soon as we account for where unemployment is expected to go over the coming months, the difference vanishes.
• After about three months, media reactions to rising and falling unemployment are statistically the same.

Figure 2:

In plain terms: the media respond equally to good and bad unemployment news.

The main takeaways from this study are the following:

1. The media are not naturally drawn to bad economic news.
   They give more attention to rising unemployment only because rising unemployment usually signals more trouble ahead.
2. Earlier research mistook the behavior of the economy for the behavior of journalists.
   Unemployment itself rises fast and falls slowly. This creates the illusion of a media bias when you only look at a single month of data.
3. When you account for where unemployment is heading, media reactions to good and bad news are the same.
4. Media coverage remains a useful source of information about the economy.
   It often reacts earlier than official statistics.
5. What looks like negativity is really the media picking up on economic momentum.
   They highlight rising unemployment because it tends to get worse — not because they prefer gloom.

To sum up

For decades, many people have believed that the media focus too much on bad economic news. Our research shows that this belief is not supported by the data.

The media’s coverage of unemployment isn’t overly negative — it’s forward-looking. Journalists report more on rising unemployment because it usually signals a bigger downturn ahead, not because they want to highlight bad news.

When economic dynamics are properly taken into account, the so-called “negativity bias” disappears.

The blog is based on “Media Multipliers: Is There a Negativity Bias in Economic News?” (With Luca Gambetti and Sarah Zoi)