When Inflation Becomes Salient: Attention Thresholds, Pricing Power, and the Uneven Economics of Corporate Margins

When Inflation Becomes Salient: Attention Thresholds, Pricing Power, and the Uneven Economics of Corporate Margins

Inflation is never only a number. At low and moderate rates it can remain a background variable, noticed by economists, central banks, and procurement departments but only intermittently by households. At higher rates it can become a social fact: people search for it, discuss it, change shopping habits, demand wage protection, shorten planning horizons, and reinterpret almost every price through an inflationary lens. The chart captures that transition. Its fitted relationship between inflation and a Google Search Index is comparatively flat below an estimated threshold of approximately 3.5%, then rises much more steeply above it. The point is not that 3.5% is a universal law. It is that public attention appears nonlinear, and that nonlinearity changes the economics of inflation itself.

The corporate side of the story is equally important. Aggregate data have sometimes shown earnings growing faster than sales during sustained inflationary periods, consistent with selling prices rising faster than wage costs. That observation can be true without implying that every company gains, that labor always loses, or that inflation mechanically widens margins. Industry structure, input contracts, inventory accounting, leverage, demand elasticity, regulation, and the timing of wage bargaining create large differences across firms and episodes. The analytical task is therefore to connect two claims without collapsing them into a slogan: inflation can cross an attention threshold, and some firms can temporarily turn that regime into operating leverage. Both mechanisms are state-dependent, heterogeneous, and eventually constrained by feedback.

Reading the chart as a regime map

The chart should be read as a piecewise description rather than a conventional straight-line correlation. Let attention be A, inflation be pi, and the threshold be pi-star. A simple representation is A = alpha + beta-low(pi) + beta-kink max(0, pi - pi-star) + epsilon. Below pi-star, the relevant slope is beta-low; above it, the slope becomes beta-low plus beta-kink. Visually, beta-low is close to flat while the post-threshold slope is strongly positive. Around the estimated 3.5% threshold, inflation stops being merely higher and begins to command disproportionate attention. This distinction matters because economic agents react not only to the level of prices but also to the visibility, persistence, and perceived controllability of price change.

Several cautions follow. Search intensity is not a pure measure of inflation expectations, hardship, or policy credibility. It can reflect media coverage, political debate, gasoline prices, housing costs, or a desire to understand unfamiliar terminology. The scatter also contains dispersion, especially around the threshold region. A fitted kink does not prove a structural break, and the exact location can move with the sample, country, search term, and inflation measure. Yet measurement imperfection does not erase the signal. Search behavior is a revealed-attention indicator: it records an active information-seeking choice. When searches accelerate nonlinearly, inflation has likely entered household decision processes more forcefully than a smooth macro model would assume.

The shape also warns against linear extrapolation. If attention barely responds while inflation moves from, say, 1.5% to 2.5%, it does not follow that a move from 3.5% to 4.5% will have the same behavioral impact. The post-threshold region may activate new channels: comparison shopping, brand substitution, wage negotiations, political pressure, inflation-indexed contracts, and faster repricing by firms. In other words, the chart is not only about searches. It is a compact map of a possible regime change in information acquisition and economic coordination.

Rational inattention and the economics of noticing

Rational-inattention theory offers a natural foundation for the pattern. Information is costly to acquire and process. Households have limited time, firms have limited managerial bandwidth, and updating every belief continuously is inefficient when the economic stakes are small. In the Sims tradition, agents optimally allocate scarce attention toward variables whose movements have sufficiently large consequences. When inflation is low and stable, a household may tolerate imprecise knowledge of the aggregate rate because the benefit from monitoring it is limited. Once inflation becomes large enough to threaten real income, financing costs, or savings, the expected value of information rises sharply.

This mechanism can generate an apparent threshold even if underlying preferences are smooth. Suppose the expected benefit of searching equals the probability that new information changes a decision multiplied by the payoff from that change. Both terms may increase with inflation. Price dispersion may widen, making comparison more valuable, while the dollar consequence of a purchasing mistake also grows. Search occurs when that product exceeds the time and cognitive cost. Aggregating across households with different costs and exposures yields a gradual but potentially steep takeoff. The estimated 3.5% point can therefore be interpreted as a population-level activation zone, not a magical boundary faced identically by every person.

The same logic applies inside companies. Management teams do not constantly redesign price architecture, compensation bands, hedging programs, and working-capital policies. But when inflation becomes salient, those activities receive senior attention. Frequency of pricing reviews increases; procurement teams diversify suppliers; treasury teams revisit duration and floating-rate exposure; boards ask whether nominal growth is masking volume deterioration. Public search activity and corporate process changes are different data, yet both can arise from the same optimization under limited attention. A macro variable that crosses the attention threshold can change the frequency with which the private sector updates decisions.

Salience, memory, and expectation formation

Behavioral economics adds another layer. Consumers overweight vivid, frequent, and personally experienced price changes. Food, fuel, rent, insurance, and utility bills are encountered repeatedly, so their movements can shape perceived inflation more than their statistical weights alone would predict. Search activity may rise when enough salient prices move together to create a coherent narrative. Once that narrative forms, people start interpreting unrelated price increases as confirmation. The result is an attention cascade: experienced inflation raises concern, concern raises information demand, and information demand makes inflation-related news more prominent.

This does not mean households are irrational. Personal inflation rates genuinely differ because consumption baskets differ. A renter renewing a lease, a commuter buying gasoline, and a retiree paying medical expenses face distinct shocks. Aggregate inflation is an imperfect summary of those experiences. Search can be a reasonable response to uncertainty about whether a personal shock is idiosyncratic or economy-wide. It may also reveal a demand for forecasting: households want to know whether today’s price increase will persist, whether wages will catch up, and whether borrowing or delaying a purchase is wiser.

Expectations research suggests that attention affects both the speed and the distribution of belief updating. In low-inflation regimes, many households update infrequently, producing sticky expectations. When inflation becomes salient, updating becomes more common, but not necessarily more accurate. People encounter noisy commentary, partisan claims, and recent-price extrapolation. Dispersion in expectations can therefore increase even as average attention rises. For monetary policy, this is a delicate state: more people are listening, but the information environment is crowded and credibility is being tested in real time.

Why the threshold can become self-reinforcing

Attention is not merely an outcome of inflation; it can become part of inflation’s propagation. If workers monitor inflation more closely, they may seek larger or more frequent wage adjustments. If firms believe customers expect price increases, managers may face less resistance to repricing. If consumers expect prices to rise further, they may bring purchases forward. Each response can be individually sensible while increasing persistence at the aggregate level. The Phillips curve is then not fixed: its short-run behavior can depend on the salience of inflation, bargaining institutions, and the credibility of the nominal anchor.

A useful distinction is between first-round and second-round effects. A commodity or supply shock directly raises selected costs. That is the first round. The second round begins when wages, service prices, rents, and broad expectations adjust in response. Search intensity is potentially informative about the transition because it indicates whether inflation has entered widespread awareness. It cannot prove that a wage-price spiral will follow, but it helps identify the environment in which propagation risks are greater. Central banks should care about the slope of attention, not only the level of headline inflation.

There is also a coordination channel. When few people attend to inflation, a firm raising prices may fear losing customers and a worker demanding a large raise may appear unrealistic. When inflation is common knowledge, both actions become easier to justify. Public narratives provide a shared explanation for nominal changes. This can accelerate adjustment and reduce relative-price confusion, which is useful, but it can also make inflation more persistent. The threshold thus marks a potential shift from isolated price changes to a coordinated nominal regime.

From nominal sales to real economics

The corporate claim requires careful decomposition. Nominal sales growth can be written approximately as price growth plus volume growth, with interaction terms omitted for simplicity. Earnings growth then depends on how revenue changes compare with changes in variable costs, wages, fixed costs, depreciation, interest, and taxes. If a company raises selling prices faster than unit labor cost and other inputs, operating profit can grow faster than revenue. That is positive margin leverage. But if price increases conceal falling volume, deteriorating mix, or future customer loss, the near-term earnings result may overstate durable economics.

Consider a stylized company with revenue of 100, wage cost of 30, other variable input cost of 40, and fixed operating cost of 20, leaving operating profit of 10 and a 10% margin. Suppose selling prices rise 8%, volumes are unchanged, wages rise 4%, other inputs rise 6%, and fixed cost rises 3%. Revenue becomes 108. Wage cost becomes 31.2, other variable cost 42.4, and fixed cost 20.6. Operating profit rises to 13.8, a 38% increase, while sales rise only 8%; the margin expands to about 12.8%. Small differences between output-price and cost inflation can create large percentage changes in profit because profit is the residual.

Now change the assumptions. If volume falls 5% because customers resist higher prices, effective revenue becomes roughly 102.6 rather than 108. If input costs were contracted at a lag and then reset upward, the apparent margin benefit can reverse. If wages catch up through annual negotiations, next year’s labor bill may rise faster than this year’s. Corporate inflation leverage is therefore a timing structure, not a permanent free lunch. The relevant question is who reprices first, how much quantity responds, and when contracts reset.

Pricing power is a distribution, not a market-wide constant

Aggregate earnings can outperform aggregate sales while many firms suffer. A sales-weighted or capitalization-weighted index may be dominated by large businesses with strong brands, concentrated market positions, global sourcing, or asset-light cost structures. Smaller firms may face the same input inflation without equivalent bargaining power. Within an industry, the leader may raise price while weaker competitors absorb costs to defend share. An aggregate result can therefore reflect composition as much as a universal causal effect.

True pricing power has several dimensions. A firm needs the ability to raise price, preserve volume, protect customer lifetime value, and avoid provoking entry or regulation. Brand strength can help, but so can switching costs, mission-critical products, scarce capacity, network effects, distribution control, and contractual escalators. Conversely, a commodity producer may enjoy high realized prices without possessing durable pricing power; it is a price taker benefiting from the cycle. Investors should distinguish cyclical price realization from structural control over price.

Industry variation is decisive. Software businesses may reprice subscriptions with low marginal cost but face renewal churn. Consumer staples can pass through costs gradually but risk private-label substitution. Industrial companies may have surcharge mechanisms yet suffer from long contract lags. Banks can benefit from higher nominal rates until deposit costs and credit losses catch up. Utilities often recover costs through regulation, but with delays. Retailers turn inventory rapidly and can reprice quickly, while construction firms may be trapped in fixed-price contracts. The phrase “companies benefit from inflation” is analytically empty unless it specifies which companies, which inflation, and which horizon.

Wage lags, labor bargaining, and the margin window

The observation that selling prices can rise faster than wages is plausible because product prices and compensation reset on different clocks. Many companies can change list prices within weeks or months. Base salaries often reset annually, collective agreements may run longer, and labor-market information arrives with delay. During an inflation acceleration, output prices may therefore jump before nominal wages fully adjust. Because wages are a major cost, that lag can temporarily transfer real income from labor to profits and produce earnings growth above sales growth.

But lagged adjustment is not the same as absent adjustment. Workers observe declining real wages, change jobs, negotiate cost-of-living clauses, or withdraw effort. Tight labor markets strengthen their response. Firms may then pay catch-up increases, retention bonuses, and higher benefit costs. The profit share can rise early in an inflation episode and compress later even if inflation itself remains elevated. Kalecki-style mark-up frameworks and modern bargaining models both emphasize distributional conflict: prices and wages are claims on nominal income, and the realized split depends on market power, institutions, and sequence.

The quality of earnings matters as well. If margin expansion comes primarily from a temporary wage lag, valuing it as a permanent increase in normalized profitability is dangerous. Analysts should estimate labor-cost sensitivity, contract renewal schedules, workforce turnover, and the gap between wage growth for job stayers and switchers. A company with low reported wage inflation may simply be postponing adjustment or losing talent. Sustainable productivity improvement is different: if output per hour rises, wages and profits can both grow without the same distributional tension.

Inventories, accounting, and the illusion of effortless leverage

Inventory accounting can amplify or obscure inflation effects. A firm that sells goods acquired at older, lower costs may report a temporary gross-margin boost when selling prices reflect current replacement costs. Under first-in, first-out accounting, this effect can be especially visible during rising input prices. Yet replenishing inventory is more expensive, so cash economics may be weaker than reported profit suggests. Last-in, first-out accounting recognizes newer costs sooner but can create other tax and comparability effects. Analysts need to reconcile income-statement margins with working-capital cash flows.

Inflation also raises the nominal amount of receivables and inventory required to support a given real volume of business. Even profitable growth can consume cash. Companies with supplier financing or negative working capital may be relatively insulated, while distributors and manufacturers can face large funding needs. Higher policy rates compound the burden. A business may report robust earnings and still experience falling free cash flow because more capital is tied up in nominal operating assets.

Depreciation creates another mismatch. Historical-cost depreciation may understate the economic cost of replacing long-lived assets when equipment and construction prices rise. Reported operating profit can look strong while the replacement-cost margin is much thinner. For capital-intensive industries, analysts should ask whether current depreciation is sufficient to maintain productive capacity in real terms. Inflation accounting is not a footnote; it determines whether apparent profit growth represents distributable surplus or merely compensation for a depreciating currency.

Demand elasticity and the limits of pass-through

The central constraint on pricing is demand elasticity. A company can raise nominal prices faster than wages only while customers retain the ability and willingness to pay. At first, nominal income growth, savings buffers, or necessity may support demand. Over time, households substitute toward cheaper products, reduce quantities, postpone durable purchases, or trade down in quality. Search attention itself may accelerate this process by making consumers more deliberate and price-sensitive.

This creates a subtle interaction between the two themes. Above the attention threshold, inflation may initially give firms social permission to raise prices, because customers understand that costs are rising everywhere. But sustained attention also teaches customers to compare, negotiate, and substitute. The same salience that facilitates early pass-through can erode later pricing power. Corporate gains are therefore most likely when firms reprice quickly, customers react slowly, and costs—especially wages—adjust with a lag. That configuration is intrinsically temporary unless supported by genuine productivity or structural market power.

Elasticity also varies across the income distribution. Higher-income households may absorb price increases, while lower-income households cut quantities sooner because necessities occupy more of their budgets. Aggregate revenue can remain resilient even as welfare losses become concentrated. A premium brand and a value retailer may both gain share, squeezing the middle. This barbell outcome helps explain why aggregate earnings sometimes look healthy during periods of widespread consumer dissatisfaction.

Market structure, markups, and the evidence problem

The relationship between inflation and markups is contested because measurement is difficult. Researchers infer markups from production functions, cost shares, financial statements, or industry data, each with assumptions. Rising profits during inflation can reflect market power, sector composition, commodity rents, fiscal support, or recovery from an earlier downturn. Conversely, stable accounting margins do not prove that firms lacked pricing power if they used it merely to offset exceptional input costs.

Corporate-finance reasoning suggests examining incremental margins rather than only levels. If revenue rises by ten units and operating profit rises by four, the incremental margin is 40%, even if the total margin remains modest. High incremental margins during inflation may indicate fixed-cost leverage or delayed cost adjustment. The diagnostic becomes stronger when combined with unit volumes, realized price, wage expense per employee, input-cost indices, and cash conversion. No single aggregate series can adjudicate the distributional story.

Competition policy also enters with lags. Concentrated industries may coordinate implicitly through public price announcements or follow-the-leader behavior without explicit collusion. Yet high margins invite entry, substitution, customer renegotiation, and political scrutiny. Regulators may impose price caps or windfall taxes in sectors viewed as essential. Market power can lengthen the margin window, but it also increases the probability of countervailing responses.

Monetary policy in an attention-sensitive economy

For central banks, the chart implies that the cost of tolerating inflation may be convex. Below the threshold, a small overshoot might have limited effects on household attention and expectations. Above it, an additional percentage point could generate disproportionately more search, discussion, and behavioral adjustment. If credibility depends partly on keeping inflation outside the zone of persistent public concern, policy should respond not only to average forecasts but also to the probability of crossing and remaining above a salience threshold.

This does not automatically justify aggressive tightening at 3.5%. Threshold estimates are uncertain, monetary policy works with long and variable lags, and supply shocks can make demand destruction a costly remedy. The correct implication is risk management. Policymakers should monitor attention measures, expectation dispersion, wage-setting frequency, price-review frequency, and pass-through alongside conventional inflation data. Communication matters more when attention is high because a larger audience is updating beliefs.

There is an asymmetry in returning to low inflation. Search attention may fall more slowly than it rose because memories persist and the price level does not reverse when the inflation rate declines. Disinflation means prices rise more slowly, not that past increases disappear. Households can remain dissatisfied even after headline inflation normalizes, while firms face resistance to further price increases and workers continue seeking real-wage restoration. The post-shock economy may therefore retain different bargaining behavior from the pre-shock economy.

Investment implications across assets and firms

Equity analysis should separate nominal growers from real compounders. A company reporting 8% sales growth in an 8% inflation environment has achieved no real top-line growth. If earnings grow 20%, investors should determine whether the difference comes from durable productivity, favorable mix, temporary wage lag, old inventory costs, or under-depreciation. Valuation multiples should be applied to normalized real cash flows, not automatically to peak nominal margins.

Balance-sheet structure changes the ranking of firms. Fixed-rate borrowers can benefit because inflation erodes the real value of liabilities, provided refinancing is distant and operating cash flow adjusts. Floating-rate or near-maturity debt can turn inflation into a financing squeeze once central banks tighten. Firms with tangible assets may have replacement-value support, but capital intensity also raises maintenance needs. Asset-light firms avoid replacement costs yet may rely heavily on labor, whose compensation catches up.

Across assets, unexpected inflation is generally unfavorable for long-duration nominal bonds because it raises discount rates and erodes fixed payments. Inflation-linked bonds provide more direct protection but remain sensitive to real yields and index construction. Equities are mixed claims: some firms pass through inflation, others are duration assets whose valuations compress. Commodities can hedge certain inflation shocks, though carry, roll yield, and starting valuation matter. A robust portfolio treats inflation as multiple regimes rather than a single factor.

Attention data may contribute to nowcasting and regime classification, but it should not become a mechanical trading signal. Search indices can be revised, seasonally distorted, manipulated by news cycles, and contaminated by correlated variables. A more defensible approach combines them with breakeven inflation, survey expectations, wage trackers, price-setting surveys, earnings-call language, and measures of price dispersion. The goal is triangulation: identify whether inflation is merely statistically elevated or has become behaviorally embedded.

A practical framework for analyzing the regime

For macro analysis, begin with level, breadth, and persistence. Is inflation concentrated in a few volatile categories or spreading through services? Are trimmed-mean measures elevated? Are wages responding, and is productivity offsetting unit labor costs? Then add attention: search intensity, news coverage, survey salience, and the frequency with which firms mention pricing. Finally assess institutions: contract indexation, bargaining arrangements, central-bank credibility, fiscal support, and competitive structure.

For company analysis, decompose revenue into price, volume, mix, and currency. Decompose cost into wages, commodities, freight, energy, rent, interest, and replacement capital. Map the reset frequency of each line. Estimate an inflation gap G = selling-price growth minus weighted unit-cost growth. A positive G supports margins today, but its duration depends on elasticity and contract lags. Stress the model by allowing volumes to fall and wages to catch up rather than assuming the current gap persists.

Management language deserves skepticism. “Pricing” can mean genuine value capture, a surcharge that merely reimburses costs, or a nominal increase accompanied by hidden volume decline. “Productivity” can mean automation, mix shift, reduced service, or delayed hiring. Compare claims with units, customer retention, market share, headcount, cash flow, and maintenance investment. Inflation rewards precise bridge analysis because nominal figures can make weak real performance look impressive.

Scenario analysis: separating the first year from the steady state

A disciplined forecast should distinguish the shock year, the catch-up year, and the steady state. In the shock year, firms with frequent repricing and slow-resetting costs can report exceptional incremental margins. In the catch-up year, wages, rents, supplier contracts, and interest expense begin to reflect the new price level, while consumers become more price sensitive. In the steady state, nominal growth may remain higher than before, but abnormal margin gains disappear unless productivity, market share, or business mix has structurally improved. Treating the shock-year income statement as an annuity is one of the most common valuation errors in an inflationary regime.

Analysts can formalize this with three scenarios. In a benign case, inflation falls without recession, wage growth converges gradually, volumes remain stable, and the firm keeps part of its price-cost advantage. In a squeeze case, volumes fall before input and wage inflation recede, producing negative operating leverage. In a persistence case, inflation remains above the attention threshold, nominal revenue stays strong, but discount rates and working-capital needs also remain high. The highest accounting earnings need not correspond to the highest equity value because the persistence case may carry a larger risk premium and lower terminal multiple.

The horizon matters for cross-sectional selection. Over a few quarters, investors may prefer firms with inventory purchased cheaply, fixed-rate liabilities, and rapid repricing. Over several years, the better businesses are those that retain customers after price resets, share productivity gains with labor, and fund replacement capital without weakening the balance sheet. Tactical inflation beneficiaries and durable compounders can overlap, but they are not identical categories. A portfolio built only around the first-order beneficiaries may be exposed when the distributional catch-up begins.

Finally, scenario probabilities should respond to attention data without being dictated by them. A sharp rise in search interest can signal that the economy has entered the persistence case, but it can also mark peak anxiety near the end of the inflation shock. Direction, breadth, and duration matter more than one observation. If searches fall while wage and service inflation remain sticky, public concern may be normalizing before the underlying process. If searches remain elevated after headline disinflation, inflation memory may continue to shape bargaining and politics. Attention is a state variable to interpret, not an oracle.

What would falsify the thesis

A serious framework must specify disconfirming evidence. The attention-threshold claim would weaken if the fitted kink disappears across reasonable samples, alternative inflation measures, and search definitions, or if search increases are fully explained by media volume without independent household behavior. It would also weaken if higher attention showed no relationship with expectation updating, consumption choices, wage bargaining, or political salience. The chart is suggestive, not sufficient by itself.

The corporate-margin claim would weaken if earnings outperformance vanished after adjusting for sector composition, inventory accounting, fiscal transfers, and base effects. It would also weaken if unit labor costs consistently rose faster than realized selling prices during the relevant periods, or if free cash flow failed to confirm accounting profit. Variation is not a nuisance to be averaged away; it is central evidence about mechanism.

Conversely, confidence would rise if microdata showed households increasing price search and substitution sharply near similar inflation levels, firms accelerating price reviews, and wage contracts resetting with predictable lags. On the corporate side, matched firm data linking market concentration, repricing frequency, wage exposure, and margin outcomes would help distinguish genuine pricing power from composition. The best evidence connects attention, action, and financial results rather than relying on any one aggregate correlation.

Conclusion: inflation changes behavior before it changes every balance sheet equally

The chart’s central message is nonlinear: below roughly 3.5%, the fitted relationship between inflation and online search attention is relatively flat; above that estimated threshold, attention rises much more sharply. The economic meaning is not a universal trigger but a change in regime. Inflation becomes worth noticing, learning about, negotiating over, and acting upon. Rational inattention, salience, heterogeneous personal experience, and coordination all help explain why the response can accelerate rather than rise smoothly.

The corporate evidence adds a second, conditional proposition. In some sustained inflationary periods, aggregate earnings can grow faster than sales because selling prices rise faster than wage costs. Profit, as the residual between revenue and cost, can respond dramatically to modest differences in repricing speed. Yet the benefit varies across industries and periods, and it may rest on temporary wage lags, old inventory costs, accounting conventions, fiscal support, or concentrated market structure. Eventually labor bargaining, customer substitution, refinancing, replacement costs, regulation, and competition push back.

Taken together, these mechanisms describe inflation as a redistribution process governed by attention and timing. Before the threshold, many agents economize on information. After it, households search, workers bargain, firms reprice, and policymakers communicate under greater scrutiny. Early movers with pricing flexibility may widen margins; late adjusters bear the cost. The investment lesson is to avoid treating nominal growth as real strength or aggregate profitability as universal pricing power. The policy lesson is that keeping inflation expectations anchored is partly about preventing a background statistic from becoming an organizing principle of everyday economic life.