Inflation & Interest
A Technical History of Inflation, Interest, and Unemployment
Even among finance practitioners the inflation–unemployment relationship is often treated as a simple, stable inverse “rule”: tighter labor markets mean hotter inflation, looser markets mean disinflation. That heuristic is useful until it isn’t. Modern macro doesn’t view the Phillips curve as a fixed law so much as a policy-contaminated, expectations-mediated object whose slope, position, and even visibility shift across regimes. The result is a gap between the surface story most people carry around and the actual intellectual lineage and empirical record that produced today’s policy framework; this piece reconstructs that lineage, and the data history, to clarify what the relationship is, and what it is not.
Why Does Unemployment Exist?
Why does unemployment exist at all? Part of the answer is mechanical. Even in a well-functioning market, matching takes time: workers separate from jobs, search, relocate, retrain, bargain, and sort. That matching friction creates a baseline flow between employment and unemployment even when the economy is not “weak.” Another part is compositional: when production structures change, some skills no longer map cleanly onto available jobs. That structural mismatch—think workers displaced by technology—can keep unemployment elevated even when vacancies exist elsewhere.
But the constraint that disciplines policy is not merely that some unemployment is unavoidable. It is that attempts to push joblessness “too low” can destabilize inflation, and that the best a central bank can do is steer unemployment toward a level consistent with stable inflation. The claim is not that policymakers prefer unemployment, but that they face a long-run constraint: unemployment tends to gravitate toward a level shaped by labor-market structure and expectations, while monetary policy is mainly responsible for nominal outcomes.
Keynes, Demand Deficiency, and the Pre-Inflationary Lens
The first major modern contribution comes with Keynes’s emphasis on involuntary unemployment. In classical theory, persistent unemployment was often blamed on wages being “too high,” perhaps due to unions or rigid norms that prevented clearing.
Keynes redirected attention to aggregate demand. A worker could be willing to work at the prevailing money wage and still fail to find a job because economy-wide spending was insufficient in the short run. If wages fell broadly, household incomes would fall too, potentially depressing demand further and worsening the deficiency. The policy implication was not a war on wages, but management of total spending to sustain full employment.
What later retellings often miss is that Keynes’s unemployment argument did not center inflation. The later integration of demand management with an inflation–unemployment trade-off came afterward. The postwar transition reframed demand management as a government obligation, largely via fiscal tools, while inflation dynamics entered the policy problem more explicitly later.
Phillips, Samuelson–Solow, and the “Menu” Interpretation
Empirical attention to the inflation–unemployment relationship is sometimes traced to Irving Fisher in the 1920s, but the object that entered policy imagination was the Phillips curve. In 1958 William Phillips documented a striking historical relationship in Britain between unemployment and wage inflation over nearly a century: low unemployment coincided with faster wage growth and high unemployment with slower wage growth. The downward-sloping curve suggested a regularity robust to deep institutional and political change.
When Samuelson and Solow examined the United States, stability was weaker; the curve appeared to shift. Still, within a given era, tight labor markets seemed associated with faster nominal wage growth, making the correlation tempting to treat as policy-relevant. The relationship became a “menu”: accept more inflation for less unemployment, or more unemployment for price stability. Whether policymakers literally believed in a menu is less important than how the framing implied the long-run constraint was preference, not structure.
Friedman and Phelps: Expectations and the Natural Rate
The decisive critique came from Friedman and Phelps. Their claim was not that monetary policy is powerless, but that the menu collapses once expectations adjust. Suppose a central bank expands money and stimulates spending to push unemployment below its sustainable level. Higher demand induces firms to hire more and raise prices; inflation rises. If workers did not anticipate inflation, real wages fall relative to what contracts assumed. Labor is temporarily cheaper in real terms, hiring increases, and unemployment drops.
But the effect is temporary. When wages are renegotiated, workers incorporate realized inflation to restore purchasing power, demanding commensurate nominal raises. If bargaining power and labor-market structure are unchanged, unemployment drifts back toward the “natural” rate: the rate consistent with stable inflation given institutions, frictions, and expectations.
Keeping unemployment below that level would require repeated inflation surprises—each time exceeding expectations. Hence the implication: sustaining unemployment below the natural rate would require inflation to accelerate persistently. This is the origin of NAIRU, the non-accelerating inflation rate of unemployment. It does not say inflation always rises when unemployment is low; it says holding unemployment below a threshold tends to push inflation upward over time, while holding it above tends to push inflation downward.
This changed the policy question. The relevant target became not literal “full employment,” but employment consistent with nominal stability. Central banks’ operational challenge shifted to inferring an unobservable natural rate from inflation and unemployment dynamics.
Stagflation and the Weakening of Naïve Stability
The 1970s seemed to vindicate the critique. Stagflation—high unemployment alongside high inflation—followed major oil-price shocks. A stable Phillips “menu” could not survive an era in which the correlation inverted or collapsed. If the curve was exploitable, the data looked damning: inflation could surge even as unemployment rose.
This did not necessarily refute the expectations-augmented view. It made it easier to argue that supply shocks and policy choices shifted the underlying natural rate. If inflation did not decelerate despite slack, proponents could claim NAIRU had moved because institutions, bargaining, productivity, or the economy’s structure had changed, and because the shock was not primarily demand-driven.
But the episode also exposed a problem: NAIRU could function as an explanatory escape hatch. When inflation and unemployment moved “wrong,” the natural rate could simply be re-estimated. The concept stayed influential because it was flexible, but that flexibility limited its value as a hard anchor.
Rational Expectations, Credibility, and the New Keynesian Turn
A second struggle concerned how expectations form. Friedman’s story leans on adaptive learning, with agents extrapolating from recent inflation. Rational expectations sharpened the critique: if firms and households anticipate policymakers’ incentives, systematic attempts to push unemployment below sustainable levels should trigger inflation quickly, not with a lag. Credibility becomes an instrument: if the central bank can credibly commit to avoiding unsustainable booms, it can anchor inflation expectations and reduce inflationary consequences of shocks.
The Volcker disinflation became the test case. Volcker tightened sharply; rates rose to extraordinary levels, recession followed, unemployment exceeded 10%, and inflation fell dramatically. One interpretation is that expectations were not effortlessly forward-looking; the public needed demonstrated pain to believe the regime shift. Another is that credibility was built precisely through willingness to impose output costs, making commitment rational to believe.
After the 1980s, the monetary consensus moved toward explicit or implicit inflation targeting. As central banks gained credibility, inflation expectations appeared more stable. Modeling shifted toward New Keynesian frameworks with forward-looking price setting and policy rules. Some versions implied a “divine coincidence”: stabilizing inflation around target would also stabilize the output gap and unemployment after shocks.
Few believed this held literally. Real labor markets contain frictions, heterogeneity, bargaining, and rigidities that break the coincidence. Still, the aspiration mattered: good inflation stabilization might also reduce employment volatility, weakening the apparent trade-off.
The Post-2008 Puzzle: Missing Disinflation and a Moving NAIRU
The global financial crisis produced a puzzle. The collapse in demand and surge in unemployment should, under conventional Phillips logic, have generated sharper disinflation. Instead, inflation fell only modestly, and in some places only briefly undershot levels consistent with price stability. If the natural rate is inferred from inflation–unemployment co-movements, many economists marked up NAIRU: perhaps unemployment had become more structural, or labor-market matching and attachment had been damaged.
But as recoveries matured, unemployment fell far below those elevated estimates without the predicted inflation surge. Estimates of the natural rate were revised down again. The whipsaw revealed a core fragility: the natural rate is not observed—it is backed out using inflation behavior as a diagnostic. If inflation is unusually inert, NAIRU becomes unusually uncertain.
Explanations multiplied. Perhaps headline unemployment undermeasures slack by excluding discouraged and marginally attached workers. Perhaps changing labor-force participation and employment composition distort the signal. Perhaps the short-run Phillips curve flattened because expectations became strongly anchored, bargaining institutions shifted, globalization and import competition constrained pass-through, or pricing power behaved differently in low-inflation regimes. Or perhaps the relationship persists locally but is hard to detect in national aggregates.
A Flattened Curve: Measurement & Rigidity
One route to flatness is mismeasurement. If low unemployment draws in workers from the periphery—people not counted as unemployed because they are not actively searching—then falling unemployment may not imply the same scarcity of labor as before. The labor market can expand on extensive margins without triggering wage pressure implied by the headline rate. Japan’s experience—expanding employment by pulling in women and older workers—illustrates how measured unemployment can be very low while latent labor supply remains.
Another route is nominal wage rigidity. Firms resist wage cuts in downturns for morale and retention reasons, but this can also make them slow to raise wages in booms—especially if increases are treated as hard to reverse. Under this view, tight markets still matter, but adjustment is sluggish and can be masked by intervening shocks.
Even if wages respond, inflation may not. Wage growth passes into prices through firms’ pricing decisions, which reflect adjustment costs and norms. In low-inflation environments, many firms reprice infrequently because changing prices is costly and strategically delicate. Menu costs are not just printing; they include customer relations, coordination, and demand risk. If prices are sticky, modest cost changes may not translate promptly into measured inflation. The economy may need to move “a lot” before aggregate prices visibly move at all. The paradox is that low, stable inflation can itself reduce inflation’s responsiveness to cyclical conditions.
Endogeneity: The Central Bank as the Hidden Hand in the Data
A subtler issue is endogeneity. The Phillips curve is observed under a policy regime that responds to inflation risks. If the central bank tightens when wage and price pressure begin to build, it prevents the inflation increase that would reveal the slope. Unemployment may rise before inflation accelerates; it may fall before inflation decelerates. The data can therefore show unemployment moving while inflation stays pinned, because policy is moving unemployment to keep inflation pinned.
The curve can be “flat” not because the mechanism vanished, but because policymakers actively lean against it. This is uncomfortable because it implies central-bank success makes the relationship harder to estimate, and therefore makes key unobservables—like NAIRU—harder to infer. Competent stabilization reduces the informational content of macro time series.
Neutral Interest Rates, Policy Space, and the Lower Bound
Here the natural rate of unemployment meets another elusive object: the neutral (natural) interest rate, the real rate consistent with inflation at target and the economy at potential. Central bankers imagine a middle zone where unemployment is neither inflationary nor disinflationary, and the policy rate can sit at neutral. The crisis-era world challenged this not only because inflation remained subdued, but because rates could not be cut without limit when inflation threatened to undershoot.
The constraint is technical and political: nominal rates cannot go far below zero without prompting substitution into cash and undermining bank deposits. When the policy rate hits the effective lower bound, conventional stabilization is constrained. In the late 2010s the Fed raised rates only modestly before pausing and reversing, underscoring that neutral rates were likely lower than previously thought. If the neutral nominal rate is low, the distance to the lower bound is small, leaving less room to respond to downturns.
Why might the neutral rate be lower? A common story emphasizes demographics and global capital flows: aging societies generate high saving relative to investment demand, pushing down equilibrium real rates. Too much desired saving chases too few investments, compressing the rate needed to clear markets. This “savings glut” narrative captures a core worry: secular forces may have moved the economy into a low-rate regime where policy is frequently constrained.
The Natural Rate: Existence Versus Knowability
At the end is a persistent tension. The natural rate of unemployment is invoked as if it were a fixed landmark. Empirically it behaves like a moving shadow, inferred from inflation dynamics that are altered by expectations, institutions, globalization, pricing frictions, and the policy regime itself reacting to the very signals economists try to read.
One can doubt point estimates without rejecting the idea. To deny the natural rate entirely requires an extreme view: either central banks cannot influence unemployment even temporarily, or they can push it arbitrarily low—even to zero—without inflation consequences. Neither is plausible as a general proposition. The more defensible position is that a long-run anchor exists structurally, but is time-varying and difficult to identify in real time.
The Modern Understanding
Despite flatness and unstable estimates, the core Friedman–Phelps discipline remains embedded in practice: there is no permanently exploitable trade-off. Attempts to hold unemployment below sustainable levels will eventually force inflation adjustment unless productivity, labor supply, or institutions shift the sustainable level itself. At the same time, the post-crisis era taught that inflation may respond slowly, expectations can remain anchored for long stretches, and chronic undershooting can be as salient a risk as runaway inflation.
This is why the current regime feels internally tense. If inflation is inert and the curve flat, policy should be patient and employment-forward. If the curve is hidden by credibility and endogeneity, complacency risks sudden unanchoring and harsh correction later. Layered on both is the neutral-rate problem: even when stimulus is appropriate, the central bank may lack rate-cutting room and must rely on imperfect tools. The combined history implies not a simple conclusion but a clearer description of what policymakers are doing: managing an economy with multiple shifting constraints—frictional labor markets, inflation shaped by expectations and rigidities, and interest rates shaped by secular forces that determine how much conventional stabilization capacity exists.
Summary of Core Claims
Unemployment persists even in good times because matching frictions and structural mismatch are intrinsic to labor markets.
The natural rate framework asserts that pushing unemployment too far below a sustainable level tends to destabilize inflation over time.
Keynes reframed short term unemployment as a demand-deficiency problem; inflation became central only in later postwar synthesis.
Phillips-type correlations tempted a menu interpretation, but Friedman–Phelps argued the trade-off is temporary once expectations adjust.
NAIRU is an inferential construct: it is backed out from inflation–unemployment dynamics and can appear to “move” as those dynamics change.
Stagflation undermined the notion of a stable exploitable Phillips curve and highlighted the role of supply shocks and shifting parameters.
Credibility and expectations formation (adaptive vs rational) are pivotal for how inflation responds to labor-market tightness.
The post-2008 era suggests the aggregate Phillips curve can look flat due to wage rigidities, pricing frictions, and/or policy endogeneity.
Central-bank reaction functions can “hide” the Phillips relationship in the data by moving unemployment to keep inflation stable.
A lower neutral interest rate compresses policy space, making the effective lower bound more binding and stabilization harder.
The natural rate likely exists structurally but is hard to identify precisely in real time, especially in low-inflation, credibility-heavy regimes.
Disclaimer
The information provided on TheLogbook (the “Substack”) is strictly for informational and educational purposes only and should not be considered as investment or financial advice. The author is not a licensed financial advisor or tax professional and is not offering any professional services through this Substack. Investing in financial markets involves substantial risk, including possible loss of principal. Past performance is not indicative of future results. The author makes no representations or warranties about the completeness, accuracy, reliability, suitability, or availability of the information provided.
This Substack may contain links to external websites not affiliated with the author, and the accuracy of information on these sites is not guaranteed. Nothing contained in this Substack constitutes a solicitation, recommendation, endorsement, or offer to buy or sell any securities or other financial instruments. Always seek the advice of a qualified financial advisor before making any investment decisions.