Skip to main content

Why does a stacked bar chart make one component look larger than it really is?

When financial news presents data broken down into multiple categories—like sector breakdown of the S&P 500, or different revenue sources for a company—the choice of chart structure (stacked bars vs. grouped bars) determines which comparisons are easy to see and which are hidden. In a stacked bar chart, components are layered on top of each other: if a company's revenue has three sources, the chart shows all three as parts of a single vertical bar, the first source at the bottom, the second in the middle, and the third on top. In a grouped bar chart, the same three revenue sources are shown as three separate bars side-by-side, making direct comparison of each source across years or companies easy. These are not neutral visual choices; stacked charts hide comparisons that grouped charts reveal, and journalists often choose the structure that makes their narrative more compelling, regardless of which structure best serves the reader's understanding. Learning to see which chart type is being used—and why—will prevent you from drawing false conclusions about the relative importance of financial categories.

Quick definition: A stacked bar chart layers multiple categories vertically (or horizontally), making it easy to see totals but hard to compare individual categories. A grouped bar chart places categories side-by-side, making individual comparisons easy but totals harder to assess.

Key takeaways​

  • Stacked bars excel at showing totals across time, but comparing individual components across periods or between entities is nearly impossible (except the bottom component).
  • Grouped bars excel at comparing individual components but make comparing totals less intuitive.
  • Financial journalists often choose stacked charts to emphasize totals or to downplay changes in individual components that would be obvious in a grouped chart.
  • A stacked bar chart showing "revenue growth" may hide the fact that a company's largest revenue source is shrinking, with growth coming only from a small, new division.
  • The "optical distortion" of stacked charts—where a component in the middle looks larger or smaller depending on what sits below it—creates visual misimpressions that grouped charts avoid.

How stacked bars hide comparisons​

Imagine a company with three revenue sources: legacy products (50% of revenue), new products (30%), and services (20%). Over a five-year period, revenue grows, but the mix changes:

Year 1: Total $100M (legacy $50M, new $30M, services $20M) Year 5: Total $150M (legacy $40M, new $80M, services $30M)

In a stacked bar chart, Year 1 appears as a $100M bar with legacy at the bottom (50% of bar), new in the middle (30%), and services on top (20%). Year 5 appears as a $150M bar with legacy at the bottom (40% of bar—now a smaller portion), new in the middle (now 53%—larger), and services on top (20%).

The stacked chart makes the total revenue growth ($100M to $150M) visually obvious: the bar got taller. But what happened to legacy revenue? It fell from $50M to $40M, a decline of 20%. This critical fact is nearly invisible in the stacked chart because legacy's starting point (the bottom, at $0) has not changed; only its height has shrunk. Your eye might not notice the change, especially if you are focused on the growing total.

Now imagine the same data in a grouped bar chart: three bars for Year 1 (legacy at $50M, new at $30M, services at $20M) and three bars for Year 5 (legacy at $40M, new at $80M, services at $30M). The decline in legacy revenue is instantly obvious—the legacy bar shrank. The growth in new products is striking—its bar more than doubled.

The grouped chart makes individual comparisons transparent; the stacked chart obscures them. A journalist who wanted to tell a story about "total revenue growth" would choose stacked. A journalist who wanted to tell a story about "the company is declining in legacy products but surging in new products" would choose grouped—or would note the category changes while presenting the total, a more honest approach.

The optical distortion of middle and upper layers​

Stacked bars create a particular visual illusion: components in the middle of a stack are harder to compare than components at the top or bottom. This is because the baseline (the starting point) for middle components changes depending on what sits below them, and human eyes struggle to compare lengths when they do not share a common baseline.

Consider three years of data for a company with three product lines:

Year 1: Total $100 (Product A $40, Product B $35, Product C $25) Year 2: Total $110 (Product A $40, Product B $40, Product C $30) Year 3: Total $120 (Product A $40, Product B $45, Product C $35)

In a stacked bar chart:

  • Product A (bottom) has a starting point of $0 each year, so its size is easy to compare ($40 stays flat).
  • Product B (middle) starts at $40 (on top of A), then $40, then $40. Its height appears to grow ($35 → $40 → $45), but your brain has to mentally subtract Product A's height to see that Product B actually grew by $10 over three years, not by a visually obvious amount.
  • Product C (top) has a moving baseline (the sum of A and B changes), so even eyeballing whether C's contribution grew or shrank is challenging.

In a grouped bar chart, all three products have the same baseline ($0) every time they are displayed, so comparing Product B across the three years is simple: the bar clearly grew from $35 to $40 to $45. Product C's growth ($25 → $30 → $35) is also obvious.

This optical distortion is why stacked charts are best used when you care about totals and individual components are secondary. If you care about how each component changes, grouped charts are superior.

When stacked charts make sense—and when they don't​

Stacked charts are honest and appropriate when:

  1. The total is the primary message. "Total company revenue grew from $100M to $150M" is the headline, and the breakdown by revenue source is secondary context.
  2. The reader needs to understand what the total is composed of, and the breakdown does not change much. If a mutual fund's holdings are 60% stocks, 30% bonds, and 10% cash every year, a stacked bar showing the composition over time is fine (and not misleading) because the stacks are essentially identical.
  3. You are combining percentages that must sum to 100%. A political chart showing the vote share of three candidates, all stacked to sum to 100%, is appropriate because the total is always 100% and the reader's intuition is that the sum matters.

Stacked charts are misleading when:

  1. The story is really about how individual components changed. If a company's legacy revenue is declining while new-product revenue is surging, a grouped bar chart reveals this story clearly; a stacked chart obscures it.
  2. Individual components have very different scales or baselines. If one component is $50M and another is $5M, the smaller one becomes invisible at the top of a stacked bar, especially if the total is $400M.
  3. The journalist is using a stacked chart specifically to hide an unpleasant trend in one category. This is where the deception occurs: presenting data in a stacked format that downplays or obscures a negative narrative (a shrinking division, a declining revenue source) while emphasizing the positive (total growth, an expanding division).

Real-world examples​

Example 1: Bank revenue and net interest margin. A financial news article on big banks publishes a stacked bar chart showing total bank revenue over ten years, divided into net interest income (NII—lending profit) and non-interest income (fees, trading, advisory). Over the decade, total revenue grew 15%. A reader sees the stacked bar grow taller and concludes the bank is thriving. But a grouped chart would reveal that NII, which was once 60% of the revenue, has shrunk to 35% as interest rates fell and lending margins compressed. Non-interest income more than doubled to offset the NII decline, a story crucial to understanding the bank's business shift. The stacked chart hides this transformation; the grouped chart would have exposed it immediately.

Example 2: Sector composition of the S&P 500. Financial media publishes stacked bar charts showing the composition of the S&P 500 over decades, with sectors color-coded. A reader sees the total (always 100%) and notices that technology's color (often bright blue or green) seems larger in recent years. But in a stacked bar where tech's starting point (baseline) has changed because other sectors shrank, the visual growth can be misleading. If tech was 5% of the index 30 years ago and is 30% today, that transformation is real and important. But if tech was 15% of the index 20 years ago and is 30% today, the doubling is less dramatic than a stacked chart might suggest (because tech's baseline grew from the bottom-up shifts in other sectors). A grouped bar showing each sector's percentage over time would make the changes clearer.

Example 3: Government revenue sources (taxes vs. borrowing). Financial news outlets publish stacked bars showing total U.S. government receipts, divided into tax revenue and borrowing. If borrowing is stacked on top of tax revenue, a reader might think borrowing "just grew by a small amount" (because it appears as a thin slice on top of the total), when in fact borrowing doubled as a component of total receipts. A grouped chart, or better yet, two separate line charts showing tax revenue and borrowing separately, would reveal the shift in fiscal composition.

Example 4: Market cap concentration in stock indices. Financial news publishes stacked bar charts showing the composition of the S&P 500 by company size (large-cap, mid-cap, small-cap) or by sector. A journalist reporting on "the dominance of mega-cap tech" might use a stacked chart that emphasizes the total index size while downplaying the fact that mega-cap tech's share has grown from 15% to 30% of the index. A grouped chart, or a line chart showing the percentage of the index each sector represents, would make this concentration shift far more obvious and would better support the headline's claim.

Example 5: Income inequality in financial reporting. An article on wealth distribution publishes a stacked bar chart showing the share of total wealth held by different income quintiles. The stacks always sum to 100%, which is appropriate (wealth is a zero-sum distribution). But if the journalist's point is that the top 1% has grown dramatically while the bottom 50% has shrunk, a grouped bar or a line chart showing each quintile's share separately would be clearer. The stacked bar obscures which quintiles are gaining and losing relative to others.

FAQ​

Q: Is one chart type always better than the other?​

A: No. Each chart type answers different questions. Stacked bars answer "what is the total?" Grouped bars answer "how do individual categories compare?" If the story requires both, you need both (e.g., separate charts, or notes acknowledging what the stacked chart obscures). Financial journalists should choose the chart type that best serves the story, not the type that makes the data look most dramatic.

Q: How do I know if a journalist chose a stacked chart intentionally to hide something?​

A: You cannot know for certain without asking. But watch for red flags: (1) a headline claiming a component is "growing" or "surging" while the stacked chart makes visual comparison of that component hard, (2) a total that is growing while components might be shrinking (a situation where grouped bars would reveal the hidden truth), and (3) a journalist who refuses to provide the underlying numbers or a grouped-bar alternative.

Q: Can a stacked chart ever fully hide the truth?​

A: Not if the underlying data is disclosed. A stacked chart without numbers is potentially misleading; a stacked chart with the exact values labeled is only misleading if the reader fails to read the labels. Always look for the numbers. If a news outlet publishes a stacked chart without values, demand the data, or find a source that provides both the visual and the numbers.

Q: Are 100% stacked bars (where all stacks sum to 100%) better or worse than absolute stacked bars?​

A: 100% stacked bars are appropriate when the total is always constant (like a portfolio allocation or a political vote split), but they can still obscure individual category changes. If your portfolio is 60% stocks, 30% bonds, and 10% cash in Year 1 and the same in Year 5, a 100% stacked bar shows no change. But if individual components changed (stocks up then down, bonds down then up, cashes shifted), the 100% stack hides those movements. You need to see the underlying numbers or use a different chart type to see those details.

Q: What is the best chart type for showing multiple categories over time?​

A: It depends on the message. For showing totals and composition (message: "here's what makes up the total"), a stacked bar is honest. For showing individual category trends (message: "each category is changing differently over time"), a grouped bar or multiple line charts are better. For maximum clarity, provide both the numbers and choose a chart type that does not obscure the story.

Q: Why do financial journalists use stacked charts at all if they hide comparisons?​

A: Stacked charts are visually compact and make totals obvious at a glance. For an article about "total bank revenue grew 15%," stacking the revenue sources into one growing bar is efficient and intuitive. The problem is when journalists use this efficiency to avoid discussing the underlying changes in components, or when editors choose stacked charts specifically to downplay unfavorable trends in categories. Awareness of what the chart type hides is your defense.

Summary​

Stacked bar charts make comparing totals easy but comparing individual components hard; grouped charts do the opposite. Financial journalists often choose stacked charts when a grouped chart would reveal uncomfortable truths (declining revenue sources, shrinking market share), exploiting the fact that middle and upper components are hard to visually assess when they do not share a common baseline. By asking for the underlying numbers and mentally converting stacked data to a grouped format (or finding a grouped chart version), you bypass the visual confusion and see the truth. Chart structure is not neutral; it is a choice that reflects what the journalist or editor wants readers to notice and what they want to hide.

Next​

→ Percentage stacked charts