What food labels, soda taxes, and grocery prices actually do to your nutrition choices

Food labels are supposed to help consumers make better decisions. But what if they’re also quietly changing prices in ways that push lower-income shoppers toward less nutritious food?

That’s one of the questions Dr. Chen Zhen is investigating with a USDA grant at the University of Georgia, where he is a professor of food economics studying consumer behavior through the lens of pricing, labeling, and nutrition policy. His work sits at the intersection of food systems, health outcomes, and economic research – and his findings don’t always confirm what policymakers and advocates expect.

He joined the Food Bullying Podcast to walk through the research on food labels, soda taxes, fruit and vegetable subsidies, food deserts, and what the data actually shows about the most effective ways to improve what Americans eat.

---

Three ways policy tries to change what people eat – and how well they work

Dr. Zhen organizes food policy interventions into three categories. Understanding the difference matters for anyone – dietitians, agribusinesses, agricultural organizations, or policymakers – trying to make sense of the food system.

Pricing strategies – the soda tax model. Taxing less nutritious foods to discourage purchase. Dr. Zhen considers this the most effective intervention for changing purchasing behavior, but notes that economists are careful about recommending taxes because they always carry consumer welfare costs. Any food tax is regressive – meaning lower-income households pay more as a percentage of their income. Political resistance is also significant and historically has derailed even well-supported efforts. Denmark implemented and then repealed a fat tax within a year despite initial unanimous cross-party support.

Access policies – removing unhealthy options or making healthy ones more accessible. The classic example is eliminating sugary drinks from schools. The research shows partial effectiveness, but with a meaningful catch: compensation. Studies have found that when students drink fewer sugary beverages at school, they tend to drink more at home. The behavior shifts context rather than changing.

Information provision – the nutrition label approach. The standard Nutrition Facts panel has two structural problems: it’s on the back of the package, where consumers often don’t look, and it requires a level of nutritional literacy to interpret that many shoppers don’t have. Serving size, in particular, is consistently misunderstood. This has driven growing interest in front-of-package interpretive labels – a single score or symbol that gives consumers an at-a-glance nutritional assessment without requiring them to do math.

---

The yogurt study – and the 40% finding that changes everything

Dr. Zhen’s research on the NuVal scoring system – a front-of-package label that was active in over 2,000 U.S. grocery stores before the company closed – produced a finding that reframes how the nutrition label conversation should work.

The study used a natural experiment: in 2014, NuVal revised the algorithm behind its scoring system, and many yogurt products received lower scores as a result. Dr. Zhen tracked what happened to sales before and after that change.

As expected, products that received lower scores saw sales decline. Products that received higher scores saw sales increase. Labels work – at least in the direction the research predicted.

But the more important finding came next. Retailers noticed the sales effects and responded by adjusting prices. When a product’s score fell and sales dropped, retailers lowered the price to bring sales back up. When Dr. Zhen accounted for that pricing response, approximately 40% of the label’s effect on consumer behavior was offset by the retailer’s pricing strategy.

In plain terms: a significant portion of what nutrition labels accomplish at the shelf is quietly undone by pricing decisions made in the back office. That number held across multiple statistical robustness checks in the yogurt category. Dr. Zhen’s current USDA grant is expanding this research to all packaged foods to determine whether the same dynamic holds more broadly.

---

The unintended consequence no one is talking about

The 40% offset finding leads to a hypothesis with significant implications for nutritional equity. When a product receives a lower nutrition score and retailers lower its price to compensate for declining sales, who responds most to that price drop?

Lower-income consumers, who are more price-sensitive than higher-income consumers.

The unintended consequence Dr. Zhen is investigating: front-of-package labels designed to steer consumers toward healthier choices may, through retailer pricing responses, actually make unhealthy foods more affordable and more attractive to exactly the population they’re meant to help. This remains a hypothesis being tested through the USDA grant – the data isn’t yet conclusive – but the mechanism is economically plausible and worth taking seriously.

---

 

What the research actually shows about soda taxes, subsidies, and food deserts

Three findings from Dr. Zhen’s published research in the American Journal of Agricultural Economics challenge some widely held assumptions about food policy:

Soda taxes do reduce consumption among low-income households. Lower-income consumers respond more strongly to soda taxes than higher-income consumers – meaning the tax achieves its behavioral goal most effectively in the demographic where sugary beverage consumption is often highest. The compensation effect (buying more elsewhere) exists but isn’t large enough to cancel out the positive dietary impact.

Fruit and vegetable subsidies alone don’t improve overall diet quality for low-income households. This was the most counterintuitive finding. When Dr. Zhen’s team simulated a fruit and vegetable subsidy, low-income households didn’t improve their overall nutrition score. In some cases, they spent the savings from cheaper produce on other less nutritious food. The subsidy moved money around without moving the needle on diet quality.

Combining taxes and subsidies works better than either alone. Taxing sugary beverages and other less nutritious foods while using that revenue to subsidize healthier options improved the Healthy Eating Index for both low-income and higher-income households. Neither tool is sufficient by itself. The combination is what produces measurable improvement.

---

The food desert research that challenges a common assumption

The concept of food deserts – geographic areas where residents lack access to nutritious food – drives significant policy attention and funding. Dr. Zhen’s review of the scientific literature on whether food deserts actually cause poor diet quality arrives at an uncomfortable conclusion: the causal evidence is weak.

The primary reason is transportation. The overwhelming majority of residents in food deserts have access to a car and already shop at grocery stores outside their immediate area. The fundamental problem isn’t access – it’s purchasing behavior inside the grocery store. People who can reach a full-service supermarket are still buying less nutritious food when they get there.

That finding doesn’t eliminate the food desert problem, but it does reframe where the most impactful interventions should focus. Getting a grocery store into a neighborhood doesn’t automatically improve what people put in their carts. Pricing, labeling, and income-based strategies may have more leverage than geographic access alone.

---

Why front-of-package interpretive labels beat taxes – in economists’ view

Dr. Zhen draws a sharp distinction between taxes and labels in economic terms. A soda tax changes behavior, but it does so by making a product more expensive – a real cost to the consumer. An effective front-of-package label changes behavior by changing preferences – shifting what consumers actually want to buy, not just what they can afford. Economists consider that distinction meaningful: preference change carries much less welfare cost than price punishment.

The NuVal system and Europe’s NutriScore are both examples of comprehensive interpretive labels that score products across the full range – not just flagging healthy items but also giving low scores to less healthy ones. Dr. Zhen prefers this approach to labels that only highlight the best performers and leave everything else unmarked. A label system that only rewards the top tier leaves consumers without information about the rest of the shelf.

---

What a food economics professor puts in his cart

Dr. Zhen has spent enough time with nutrition facts panels that he now understands them – but he was in graduate school before he figured out what a serving size actually meant. He steers his children away from packaged foods where possible, and when they do choose packaged items, he uses interpretive labels to identify the healthier version within a category. He buys on sale, watches prices, and tries to push more fruits and vegetables into his family’s diet with mixed success – the same challenge most parents face regardless of how much they know about nutrition science.

---

Connect with Dr. Chen Zhen: Search “Chen Zhen University of Georgia” to find his faculty page and contact information directly.

Want to bring evidence-based food information and consumer trust research to your next agricultural or dietitian event? Michele Payn speaks to agricultural organizations, agribusinesses, and dietitian associations on food bullying, consumer decision-making, and the science behind modern food production. Book Michele to speak →