Figure 1: The Trend Unemployment Rate and Its Distribution by State and
Duration
Panel A. Total Trend Rat
Panel B. Distribution by State
Panel C. Short-term Component Panel D. Long-term Component
Notes to figure: Panel A displays the
aggregate unemployment rate (black line) and the trend U-star (blue
line), with the 68% posterior interval shown as a light-blue shaded
region. Panel B shows the total trend unemployment rates for the 50
states and the District of Columbia (thin lines), along with the
labor-force-weighted mean (solid line) and median (dashed line); the
grey shaded band represents the 90% cross-state distribution. Panels C
and D present analogous figures for the short-term and long-term
components—defined as unemployment spells of six months or less and
longer than six months, respectively. The vertical shaded bars denote
periods of recession as defined by the NBER. Source:
Authors’ calculation.
DESCRIPTION: A four-panel figure presenting aggregate and state-level estimates of trend U-star alongside its short-term and long-term components. VISUAL SUMMARY: Panel A (Total Trend Rate): The actual unemployment rate (black) fluctuates sharply—peaking near 11% in the early-1980s recession and again near 10% during the Great Recession—while the trend U-star (solid blue) provides a smooth envelope around 6.5–7% in 1978, declining gradually through the 1980s and 1990s to about 5%, edging up slightly between 2000 and 2010, and then resuming its decline to roughly 4% by 2025. The 68% posterior credible interval (light-blue shading) is narrow, indicating high precision in the aggregate estimate. NBER recession bands are visible as grey vertical shading throughout. Panel B (Distribution by State): Fifty-one thin light-blue spaghetti lines fan out across the chart, spanning roughly 1–10% at various points. The weighted mean (solid black) and median (dashed black) track each other closely, starting near 6.5–7% in 1978 and declining to about 3.5–4% by 2025. The grey 90% cross-state band is widest in the early 1980s and narrows appreciably by the 2000s, reflecting convergence in short-term components across states. Panel C (Short-term Component): A qualitatively similar spaghetti chart but shifted lower. State-level short-term U-stars start around 5–6% and decline to roughly 2.5–3% by 2025. The mean/median lines follow this downward trajectory with less volatility, and cross-state dispersion narrows noticeably over the sample. Panel D (Long-term Component): Values are much smaller (y-axis: 0–2%). The mean and median lines are nearly flat, hovering around 0.7–1% through the 1990s, edging slightly upward to about 1–1.1% between 2000 and 2010, and then declining modestly thereafter. Cross-state dispersion is small and remains roughly constant, in sharp contrast to the convergence seen in the short-term panel.
Figure 2: Overall changes in trend U-star by state
Notes to figure: This figure displays changes
in trend U-star between the average values in 1978 and those in the
five-quarter period 2024:Q1–2025:Q1. The red dashed line is the
corresponding national average. Source: Authors’
calculation.
DESCRIPTION: A bar chart showing the change in each state's trend U-star from its 1978 average to the 2024:Q1–2025:Q1 average, sorted in descending order. VISUAL SUMMARY: The chart is sorted from the state with the largest decline on the left to the smallest (or negative) on the right. The y-axis ("Changes") runs from about -2 to +6 percentage points. The vast majority of states show positive values, meaning trend U-star fell over the sample period. The red dashed horizontal line (national average, approximately 2.7–2.8 pp) divides the chart: roughly twenty states to the left show larger declines, while about twenty-five states to the right show smaller declines. Three to four states at the far right have bars near zero or just below it, indicating essentially no decline or a slight increase. The bar heights form a smooth descending staircase, with the tallest bars reaching 4–4.5 pp at the top of the left end.
Figure 3: Evolution of Cross-State Dispersion in Trend Unemployment
Panel A. Total
Panel B. Short-term Component
Panel C. Long-term Component
Notes to figure: The panels show the standard
deviation (blue line) and interquartile range (IQR) of the state-level
trend unemployment rate estimates at each point in time. Panel A uses
the estimates of total trend unemployment, while Panels B and C use the
short- and long-term components, respectively. Source:
Authors’ calculation.
DESCRIPTION: A three-panel figure showing the time path of two dispersion statistics—the standard deviation (blue line) and the interquartile range (orange/red line)—for the cross-state distribution of trend unemployment. VISUAL SUMMARY: Panel A (Total): Both the standard deviation (blue) and IQR (orange) begin around 0.15–0.25 percentage points in the early 1980s, peak slightly in the mid-1980s, then decline steadily through the mid-2000s to around 0.09–0.13 percentage points, before stabilizing and remaining relatively flat through 2025. Panel B (Short-term Component): Near-identical pattern to Panel A—the two dispersion measures start around 0.13–0.22 pp in the early 1980s and decline to 0.06–0.08 pp by the mid-2000s, then flatten. This mirrors the total-trend convergence, confirming that the short-term component is the primary driver of overall cross-state convergence. Panel C (Long-term Component): A distinctly different picture. Both measures remain persistently low and nearly flat throughout the entire sample, with the standard deviation around 0.04–0.05 pp and the IQR around 0.04–0.07 pp. There is a small, gradual hump around 2008–2012 (the Great Recession period), but no secular decline. Cross-state heterogeneity in the long-term component has not converged.
Figure 4: Average Level of Trend U-Star and the Relative Importance
of Its Long-Term Component Across States
Panel A: Average Level of Trend U-star (1978:Q1-2025:Q1, %)
Panel B: Long-term Share (1978:Q1-2025:Q1, %)
Note to figure: Panel A displays the average
level of trend U-star for U.S. states (in %). Panel B displays the share
of each state’s total trend U-star that is accounted for by its
long-term component (also in %). The sample period is
1978:Q1-2025:Q1.
Source: Authors’ calculation.
DESCRIPTION: Two bar charts (sorted by magnitude) covering the full sample 1978:Q1–2025:Q1. VISUAL SUMMARY: Panel A (Average Level of Trend U-star): States are sorted from highest to lowest average trend U-star. Alaska (AK) anchors the left end at ~7.5%, followed closely by DC and WV. The bars decline in a smooth staircase to Nebraska (NE) at the far right at ~3.25%. The middle of the distribution clusters between 4.5–6%. The range between the extreme states spans more than 4 percentage points. Panel B (Long-term Share): States are independently sorted from highest to lowest long-term share. DC leads at roughly 25–26%, followed by NJ and CT in the high-20s. The bars decline gradually to Wyoming (WY) at the far right near 10%. Most states fall between 13–22%. The ranking of states by long-term share differs from the ranking by U-star level, but high-U-star states tend to also cluster toward the left in this panel.
Figure 5: Trend U-Star Levels and Long-term Shares
Panel A: Average Level of Trend U-star
Panel B: Long-term Share
Note to figure: This figure displays U-star
quartile groupings (Panel A) and the long-term share groupings (Panel B)
across states. The left column shows states in the lowest quartile (“1st
Q”); the center column shows the middle two quartiles, and the right
figure shows the highest quartile (“4th Q”). Colors correspond to the
scales shown on the right. The sample period is
1978:Q1-2025:Q1.
Source: Authors’ calculation.
DESCRIPTION: Two sets of three U.S. choropleth maps each, grouping states by quartile of average U-star level (Panel A) and by quartile of long-term share (Panel B). VISUAL SUMMARY: Panel A (U-star level): The left map (1st quartile, lowest U-star) shows a cluster of dark-purple states in the Upper Midwest and Great Plains—Nebraska, North Dakota, South Dakota, Iowa, Montana, and a few adjacent states. The center map (middle two quartiles) spreads across many regions in medium pink/purple tones, covering much of the South, mid-Atlantic, and Mountain West. The right map (4th quartile, highest U-star) highlights states in yellow/orange, including industrial Midwest states (Ohio, Michigan), several Deep South states, Alaska, and California. The color scale runs from dark purple (~3.5%) to bright yellow (~7%). Panel B (Long-term share): The left map (lowest LT share) in dark purple again features Great Plains and Mountain states (Wyoming, Nebraska, the Dakotas, Montana). The center map fills in states across most regions. The right map (highest LT share) in yellow/orange features Northeastern and Mid-Atlantic states (NJ, DE), Rust Belt states (IL, WV), and DC. The color scale runs from dark purple (~14%) to bright yellow (~25%). Comparing Panel A to Panel B: states with high U-star tend also to have high long-term shares, and states with low U-star tend to have low long-term shares, consistent with the positive association documented in Figure 6.
Figure 6: Trend U-Star and Its Long-term Share
Panel A. All States
Panel B. By Census Region
Notes to figure: The charts plot the share of
U-star that is accounted for by its long-term component (\(y\)-axis) against the average level of
U-star (\(x\)-axis) for each state. The
black lines in panel B are OLS regression lines.
Source: Authors’ calculation.
DESCRIPTION: Two scatterplot panels, with each dot representing one state: average U-star on the x-axis and long-term share (%) on the y-axis. Panel B breaks this out by Census region with OLS lines. VISUAL SUMMARY: Panel A (All States): The scatter shows a clear upward-sloping cloud with considerable dispersion. The x-axis spans from ~3% (NE, SD, ND at the lower-left) to ~7.5% (AK at the far right). The y-axis spans ~10% (WY at the bottom) to ~26% (DC at the top). States are color-coded by Census region: blue for South, orange for West, yellow for Northeast, and purple for Midwest. Notable outliers include DE (very high LT share for a moderate U-star) and AK (very high U-star but only a moderate LT share of ~14%). The overall positive slope is clear despite dispersion. Panel B (By Census Region): Four sub-panels with black OLS lines. - South (blue dots): positive slope, moderate R². States spread along a gentle upward line. - West (orange dots): positive slope, but with more scatter and a shallower line. AK pulls the right end of the x-axis far out. - Northeast (yellow dots): steep positive slope with relatively tight cluster. States like NJ and CT sit at high LT shares even at moderate U-star levels; NH and VT sit at the low end. - Midwest (purple dots): steepest slope and tightest fit (highest R²). Low-U-star states like NE, ND, SD have very low LT shares; high-U-star states like IL, OH, MI have very high LT shares. The relationship is almost linear here. --- LANDSCAPE FIGURES (Federal Reserve Districts) --- Each district figure uses a consistent two-panel layout per state: the left panel shows the state's unemployment rate (black line) and trend U-star (red line) with 68% posterior grey band; the right panel shows the duration decomposition of trend U-star into four components—1 month (red), 2–3 months (dashed magenta), 4–6 months (dotted blue), and 7+ months (solid black)—with NBER recession shading throughout.
Figure 7: The First District
(a) Connecticut
(b) Maine
(c) Massachusetts
(d) New Hampshire
(e) Rhode Island
(f) Vermont
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Six-state, 3×2 grid of dual-panel plots for New England states. VISUAL SUMMARY: All six states exhibit a clear declining trend in U-star (red line), but the starting levels and pace of decline vary. Connecticut and Rhode Island start relatively high (~6–7%) while New Hampshire and Vermont start lower (~4.5–5%). Maine begins around 7–8% and shows the steepest overall decline. The actual unemployment rate (black) shows sharp cyclical spikes around the early 1980s recession, the early 1990s recession, and the Great Recession. In the duration decomposition panels, the 1-month component (red) is the largest and shows the clearest declining trend across all six states, confirming that declining short-term inflow rates drive most of U-star's fall. The 7+ month component (black) rises noticeably after the Great Recession in all states—particularly in Rhode Island and Connecticut—before declining modestly. The 4–6 month component (dotted blue) remains small and roughly flat throughout. New Hampshire has the lowest overall trend U-star in this group and the smallest long-term component.
Figure 8: The Second District
(a) New York
(b) New Jersey
(c) Connecticut
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Three-state plot for the New York Fed district. VISUAL SUMMARY: New York has a high starting trend U-star (~7.5–8%) that declines to around 4.5% by the end of the sample; the actual unemployment rate shows two prominent peaks (early 1980s and Great Recession), with the 2008–2010 spike reaching roughly 9–10%. New Jersey starts similarly high (~7–8%) and declines to ~5%; its long-term component (black in the duration panel) rises substantially after 2008 and remains elevated, reflecting structural unemployment that built up during and after the Great Recession. Connecticut's chart (repeated from Figure 7) shows a similar pattern. In all three states, the 1-month component is dominant but clearly declining, while the 7+ month component rises persistently after 2010.
Figure 9: The Third District
(a) Pennsylvania
(b) Delaware
(c) New Jersey
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Three-state plot for the Philadelphia Fed district. VISUAL SUMMARY: Pennsylvania starts with a high trend U-star (~6.5–7%) and declines to ~5%. Its actual unemployment rate peaked very sharply in the early 1980s (~12–13%), reflecting heavy manufacturing exposure. Delaware starts around 5–6%, with a notable hump in its actual rate during the Great Recession; its long-term component rises clearly after 2008. In both states, the 7+ month duration component rises after the Great Recession and converges toward or exceeds the 2–3 month component by the end of the sample. New Jersey is repeated from Figure 8.
Figure 10: The Fourth District
(a) Ohio
(b) Kentucky
(c) West Virginia
(d) Pennsylvania
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Four-state, 2×2 grid for the Cleveland Fed district. VISUAL SUMMARY: These are among the more structurally distressed states in the paper. Ohio's trend U-star starts near 8–9% and declines to ~5%, with sharp actual-rate peaks in recessions. Its long-term component (black) rises markedly after 2008, approaching 1.5 pp by around 2015 before declining. Kentucky starts very high (~9–10%) and shows a substantial long-term component rise post-2010. West Virginia is the most dramatic: actual unemployment starts near 10–12% in the early 1980s, declines significantly, but its trend remains elevated (~6–7%), and there is a notable spike during the COVID period. The long-term component of West Virginia is large and shows a distinct rise-and-fall hump peaking around 2013–2015. Pennsylvania (repeated) is relatively calmer by comparison.
Figure 11: The Fifth District
(a) Virginia
(b) Maryland
(c) North Carolina
(d) South Carolina
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Four-state, 2×2 grid for the Richmond Fed district. VISUAL SUMMARY: Virginia has one of the smoothest trend U-stars in this group, starting around 5–6% and declining to 3.5–4%. Its long-term component remains low throughout and barely rises post-Great Recession—a reflection of Virginia's large government/defense employment base. Maryland is similar but starts slightly higher. North Carolina and South Carolina both start around 5–7% and exhibit a distinctive rise in their trend U-stars during the 2000–2010 period (visible as a bump in the red trend line), driven by manufacturing job losses and rising long-term unemployment. The 7+ month component for both Carolinas rises noticeably after 2008 and remains elevated relative to earlier decades, indicating persistent structural unemployment.
Figure 12: The Sixth District
(a) Alabama
(b) Florida
(c) Georgia
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Three-state plot for the Atlanta Fed district. VISUAL SUMMARY: Alabama starts with a high trend U-star (~7–8%) and declines to ~5%; its long-term component rises post-Great Recession. Florida shows the most dramatic cyclical behavior in this group: the actual unemployment rate swings sharply—including a very sharp spike during the Great Recession (reaching 11–12%)—while the trend U-star remains smoother, declining from ~6.5% to ~5%. Florida's long-term component rises substantially after 2008. Georgia starts around 5.5–6%, and shows a distinct bulge in the trend around 2005–2012 driven by a rising long-term component, before declining. In the duration panels for all three states, the 7+ month component (black) shows a pronounced hump peaking around 2012–2013 before declining.
Figure 13: The Seventh District
(a) Iowa
(b) Illinois
(c) Indiana
(d) Wisconsin
(e) Michigan
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Five-state plot (3×1 plus 2×1 grid) for the Chicago Fed district. VISUAL SUMMARY: Iowa is the standout low-U-star state in this group: its trend starts around 5% and declines to 3.5%, with very smooth inflows and a negligible long-term component that shows little post-recession increase. Illinois and Indiana both start higher (~7–8% and ~6–7%, respectively) with more volatility, and both show clear rises in their long-term components post-Great Recession. Wisconsin is intermediate, starting around 5–6% and showing moderate long-term component growth after 2008. Michigan is the most dramatic: the actual unemployment rate spikes to roughly 15% during the Great Recession—the highest peak of any state in the paper— and the trend U-star starts near 7–8%. Michigan's duration panel (with a y-axis extending to 5%) shows a dramatic rise in the 7+ month component after 2008, peaking near 2 pp around 2010–2012, before declining.
Figure 14: The Eighth District
(a) Missouri
(b) Kentucky
(c) Mississippi
(d) Arkansas
(e) Illinois
(f) Tennessee
(g) Indiana
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Seven-state, 4×2 grid (with one state in a final row) for the St. Louis Fed district. VISUAL SUMMARY: This is the largest district figure and shows considerable internal heterogeneity. Mississippi and Arkansas stand out with very high starting trend U-stars (~8–9%), reflecting structural challenges in these labor markets. Mississippi's long-term component shows a large and persistent rise after 2008, reaching close to 2 pp before declining. Tennessee starts at about 6–7% and shows a clear hump in the trend during 2005–2015 driven by rising long-term unemployment. Missouri starts around 6% and declines to about 4.5–5%. Kentucky (repeated from Figure 10) shows its large long-term component rise again, while Illinois and Indiana (repeated from Figure 13) display moderate long-term component growth. Across all seven states, the 1-month component (red) dominates the duration panels and generally trends downward, while the 7+ month component (black) shows a clear cyclical hump peaking around 2012–2014.
Figure 15: The Ninth District
(a) Minnesota
(b) Montana
(c) North Dakota
(d) South Dakota
(e) Wisconsin
(f) Michigan
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Six-state, 3×2 grid for the Minneapolis Fed district. VISUAL SUMMARY: North Dakota and South Dakota are among the lowest-U-star states in the entire paper. Both trend lines start around 4–5% and decline to about 3–3.5%. Their actual unemployment rates are also remarkably stable, rarely deviating far from the trend. The duration panels for both states show a minimal long-term component (the black 7+ month line barely rises above zero), confirming the very low structural unemployment in these states. Montana starts slightly higher (~6%) and shows moderate long-term component growth post-2008. Minnesota starts around 4.5–5% with a smooth decline. Wisconsin and Michigan (repeated from Figure 13) are visible again— Michigan's dramatic Great Recession spike stands out clearly compared to the other calmer states in this district.
Figure 16: The Tenth District
(a) Oklahoma
(b) Kansas
(c) Nebraska
(d) Missouri
(e) Colorado
(f) Wyoming
(g) New Mexico
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Seven-state, 4×2 grid (with one state in a final row) for the Kansas City Fed district. VISUAL SUMMARY: Nebraska is the standout in this district and arguably in the entire paper: its trend U-star starts around 3.5–4% and declines to just above 2.5%, with a smooth trend line hugging the actual unemployment rate closely. Nebraska's duration panel is almost entirely dominated by the 1-month component, with negligible contributions from longer durations. Wyoming similarly has very low trend U-star and a minimal long-term component. Kansas starts around 4.5% with moderate decline. Oklahoma exhibits somewhat higher volatility in the actual unemployment rate, reflecting the oil sector, and shows a modest long-term component rise post-2010. Colorado has larger cyclical swings in the actual rate (e.g., the dot-com bust in 2001 and the Great Recession are clearly visible) while the trend U-star remains smoother at around 4.5–5%. New Mexico shows considerable volatility in the actual UR, likely reflecting oil and gas exposure. Missouri is repeated from Figure 14.
Figure 17: The Eleventh District
(a) Texas
(b) Louisiana
(c) New Mexico
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Three-state plot for the Dallas Fed district. VISUAL SUMMARY: Texas has a relatively smooth trend U-star starting around 6% and declining to about 4%. The actual unemployment rate is moderate and shows the usual recession spikes. Texas's long-term component rises modestly after 2008 but remains small. Louisiana starts with a high trend (~7–8%) that declines, but retains a substantial long-term component that rises noticeably post-Great Recession, reaching nearly 1.5 pp by the mid-2010s—consistent with the structural challenges in that state's labor market. The duration panel's long-term line (black) shows a distinctive rise and plateau. New Mexico (repeated from Figure 16) shows its usual high volatility.
Figure 18: The Twelfth District
(a) California
(b) Arizona
(c) Utah
(d) Oregon
(e) Washington
(f) Idaho
(g) Nevada
(h) Hawaii
(i) Alaska
Notes to figure: For each state,
the left panel displays the trend U-star (red line) and its 68 percent
posterior interval (grey area) along with the state’ unemployment rate
(black line). The right panel displays the duration components of the
trend U-star: 1 month (red); 2-3 months (dashed magenta); 4-6 months
(dotted blue); 7+ months (black). Authors’ calculation.
DESCRIPTION: Nine-state, 5×2 grid (with one state in a final row) for the San Francisco Fed district. VISUAL SUMMARY: This is the largest landscape figure and spans nine states with highly diverse labor markets. Alaska anchors the figure with the highest trend U-star of any state in the paper— starting around 8–9% and declining only to about 6–7% by 2025; the duration panel shows all four components remain relatively large throughout, with the 1-month component (red) clearly dominant but declining. Nevada exhibits the most dramatic cyclical behavior: the actual unemployment rate spikes to roughly 13–14% during the Great Recession (the second-highest spike in the paper after Michigan), and the long-term component of trend U-star rises sharply post-2008. California starts around 7.5–8% and declines to roughly 5.5–6%; its long-term component rises noticeably after the Great Recession. Arizona shows a very sharp actual-UR spike during the Great Recession and a clear hump in the long-term component trend afterward. Oregon has a high starting U-star (~6–7%) with a moderate long-term rise post-2008. Washington starts high (~6.5–7%) but declines to about 4%, reflecting the technology sector's stabilizing influence; the trend is smoother than many other states. Utah and Idaho have lower and faster-declining trend U-stars (reaching 3–4% by the end of the sample) with modest long-term components. Hawaii shows a distinctive large spike in the actual unemployment rate during the COVID-19 period (likely due to its tourism-dependent economy) while the trend is smoother and lower (~4–5%).
Figure B.1: Factor estimates (1)
(a) Alaska
(b) Alabama
(c) Arkansas
(d) Arizona
(e) California
(f) Colorado
(g) Connecticut
(h) District of Columbia
(i) Delaware
(j) Florida
Notes to Figure B.1:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Ten-state appendix figure covering states A–F alphabetically. VISUAL SUMMARY: Alaska (AK) shows moderate-amplitude inflows and relatively stable parameters, but with a notable sharp drop in β₂ (curvature) post-Great Recession, indicating growing duration dependence. Alabama (AL) has a gently declining inflow trend with a clear drop in β₀ (level). Arkansas (AR) shows declining inflows and a distinctive U-shape in β₁ (slope). Arizona (AZ) has very volatile inflows—large recession spikes and a strong declining trend in β₀. California (CA) displays large cyclical swings in inflows and a steep secular decline in β₀; its β₂ (curvature) drops sharply post-2008, reflecting California's pronounced long-term unemployment surge. Colorado (CO) shows moderate inflow volatility. Connecticut (CT) has relatively smooth inflows and a clearly declining β₀. DC has the highest inflows in this figure but a smooth trend, and its β₂ drops notably after the Great Recession. Delaware (DE) shows clearly rising β₂ (becoming more negative) post-2008. Florida (FL) has highly volatile inflows (cyclically very sensitive) and a dramatic β₂ drop post-2008—one of the more striking patterns in this figure.
Figure B.2: Factor estimates (2)
(a) Georgia
(b) Hawaii
(c) Iowa
(d) Idaho
(e) Illinois
(f) Indiana
(g) Kansas
(h) Kentucky
(i) Louisiana
(j) Massachusetts
Notes to Figure B.2:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Ten-state appendix figure covering states G–M. VISUAL SUMMARY: Georgia (GA) shows large cyclical swings in inflows and a pronounced β₂ decline post-Great Recession. Hawaii (HI) has relatively stable inflows with moderate β₀ decline; the curvature is distinctive with a tight posterior band. Iowa (IA) is one of the more visually calm states: nearly flat inflows with minimal trend, stable β₀, and a β₂ curvature that barely moves post-recession—consistent with Iowa's stable agricultural/manufacturing labor market. Idaho (ID) has more volatile inflows and a clearer β₀ decline. Illinois (IL) and Indiana (IN) both show clear inflow declines and sharp β₂ drops post-Great Recession. Kansas (KS) has smooth, low-level inflows. Kentucky (KY) shows declining inflows and a clear β₂ structural shift post-2008. Louisiana (LA) has volatile inflows and a β₂ profile reflecting its oil-sensitive labor market. Massachusetts (MA) shows notably smooth inflows (tech/services economy) and a steady β₀ decline.
Figure B.3: Factor estimates (3)
(a) Maryland
(b) Maine
(c) Michigan
(d) Minnesota
(e) Missouri
(f) Mississippi
(g) Montana
(h) North Carolina
(i) North Dakota
(j) Nebraska
Notes to Figure B.3:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Ten-state appendix figure. VISUAL SUMMARY: Michigan (MI) is the most dramatic state in this figure: inflows spike enormously during the Great Recession (the largest spike in the appendix), and the β₂ curvature shows the sharpest post-2008 decline of any state—consistent with Michigan's extreme long-term unemployment surge. Minnesota (MN) and Montana (MT) show calmer patterns. Mississippi (MS) has elevated inflows and a clear β₂ decline post-2008, reflecting its persistently high structural unemployment. Missouri (MO) is moderate in all parameters. North Dakota (ND) is one of the visually flattest states: inflows are nearly horizontal, β₀ is stable, and β₂ barely moves post-recession—confirming that North Dakota's labor market has been largely unaffected by the structural shifts seen elsewhere. Nebraska (NE) shows similarly flat inflows with a very stable β₁ slope. In sharp contrast, North Carolina (NC) has volatile inflows and a pronounced β₂ drop post-Great Recession. Maryland (MD) and Maine (ME) show moderate parameter dynamics typical of Northeast states.
Figure B.4: Factor estimates (4)
(a) New Hampshire
(b) New Jersey
(c) New Mexico
(d) Nevada
(e) New York
(f) Ohio
(g) Oklahoma
(h) Oregon
(i) Pennsylvania
(j) Rhode Island
Notes to Figure B.4:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Ten-state appendix figure. VISUAL SUMMARY: Nevada (NV) stands out with the most volatile inflows in this figure and a dramatic β₂ curvature drop post-2008—the second most dramatic in the appendix after Michigan— reflecting Nevada's extreme long-term unemployment surge. New Hampshire (NH) is quiet and stable in all parameters. New Jersey (NJ) shows declining inflows and a pronounced β₂ drop. New Mexico (NM) has volatile inflows and a relatively flat but downward-sloping β₁ parameter with a nearly horizontal β₂. New York (NY) shows declining inflows and a clear β₂ post-Great Recession drop. Ohio (OH) displays declining inflows and a sharp β₂ structural shift. Oklahoma (OK) shows moderate inflow volatility with oil-sector cyclicality. Oregon (OR) has moderately volatile inflows and a clear β₂ decline post-2008. Pennsylvania (PA) shows declining inflows and gradual parameter shifts. Rhode Island (RI) is one of the most structurally affected small states, with a prominent β₂ curvature drop post-Great Recession visible even on the small chart.
Figure B.5: Factor estimates (5)
(a) South Carolina
(b) South Dakota
(c) Tennessee
(d) Texas
(e) Utah
(f) Virginia
(g) Vermont
(h) Washington
(i) Wisconsin
(j) West Virginia
Notes to Figure B.5:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Ten-state appendix figure. VISUAL SUMMARY: South Dakota (SD) closely resembles North Dakota from Figure B.3: nearly flat inflows, stable β₀ level, and minimal β₂ movement—another example of a structurally stable agricultural labor market. South Carolina (SC) shows more pronounced inflow volatility and a clear β₂ drop post-2008. Tennessee (TN) has declining inflows and notable β₂ curvature shift. Texas (TX) shows large inflows (high absolute level) but a smooth declining trend, and moderate post-recession β₂ movement. Utah (UT) has high cyclical inflow volatility but a clearly declining β₀ trend. Virginia (VA) shows very stable and gently declining inflows, consistent with its large government employment share; β₂ barely moves post-recession. Vermont (VT) is calm and stable. Washington (WA) shows gradually declining inflows and a β₂ profile that is more stable than most coastal states, likely reflecting the technology sector's resilience. Wisconsin (WI) and West Virginia (WV) are both visible; WV has a distinctive β₂ curvature that drops sharply post-2008 and has wide posterior bands, consistent with its elevated long-term structural unemployment.
Figure B.6: Factor estimates (6)
(a) Wyoming
Notes to Figure B.6:
For each state, the inflows are displayed in the upper left panel, \(\beta_0\) in the upper right panel, \(\beta_1\) in the lower left panel, \(\beta_2\) in the lower right panel. The
black lines are the factor estimates, the solid blue lines are the
posterior means of trend estimates, and the dashed blue lines are the 68
percent posterior intervals. Shaded areas are NBER recessions.
Source: Authors’ calculation.
DESCRIPTION: Single-state appendix figure for Wyoming. VISUAL SUMMARY: Wyoming occupies a single 2×2 grid on an otherwise blank page. Its inflows are moderate and relatively flat, with some cyclical variation (oil/gas exposure is apparent as moderate spikes). The β₀ (level) shows a slow declining trend with a somewhat wide posterior band. The β₁ (slope) is negative and shows a gradual steepening. Most notably, Wyoming's β₂ (curvature) is nearly flat through most of the sample and shows only a mild post-Great Recession decline with wide 68% posterior bands—consistent with Wyoming having the lowest long-term share of trend U-star of any state in the paper (roughly 10%). The posterior uncertainty is visually larger for Wyoming than for most states in earlier figures, reflecting the smaller sample size and lower informational content in this state's relatively stable labor market.