Accessible Version
Population Growth and Labor Market Fragility: Lessons from Domestic and International Experiences, Accessible Data
Figure 1. Labor Force and Payroll Employment before the Average Recession
This is a figure with two bar charts side-by-side showing data for the average experience over the year prior to the four recessions included (1980, 1990, 2001, 2007).
The left panel is titled “A. Average annual labor force and employment growth in year prior to recession.” The panel displays the annual percent change in labor force and payroll employment comparing the slowest (black bars) versus fastest (red bars) population growth states. For labor force growth (left-most set of bars), slow-growth states show approximately 0.5 percent growth while fast-growth states show approximately 1.7 percent growth. For payroll employment (right-most set of bars), slow-growth states show approximately 0.7 percent growth while fast-growth states show approximately 2.5 percent growth.
The right panel is titled “B. Share of months with negative employment growth a year prior to recession.” The panel shows the share of months with negative payroll employment growth, with slow-growth states (black bar, left) experiencing negative growth in approximately 40 percent of months compared to approximately 20 percent for fast-growth states (red bar, right).
Note: Key identifies in order from left to right. Labor force from BLS’s Local Area Unemployment Statistics and employment from BLS’s Current Employment Statistics data.
Source: BLS data via Haver Analytics, Authors’ calculations.
Figure 2. Unemployment Rates in the Average Recession
This is a line chart showing average monthly data from 12 months before to 48 months after the start of the recession (including the 1980, 1990, 2001, and 2007 recessions). The y-axis displays the deviation of the unemployment rate relative to its peak (in month 0, denoted by the vertical dashed line), ranging from approximately –0.5 to 3.5 percentage points. There are two lines: the black dashed line represents the unemployment rate deviation relative to peak for the states in the slowest quintile of population growth prior to the recession and the red line is for the fastest quintile of population growth. Both lines follow nearly identical paths, starting at around -0.25 percent before the recession, then rising sharply to peak at approximately 3 percentage points around month 24, before gradually declining to 2 percentage points at 48 months. Despite the stark differences in population growth rates between these groups, their unemployment rate patterns during recessions are remarkably similar.
Note: Unemployment from BLS’s Local Area Unemployment Statistics.
Source: BLS data via Haver Analytics, Authors’ calculations.
Figure 3. Sahm Rules at the Aggregate and State Level
This is a line chart showing monthly data from 1979 to 2025. The shaded regions are months when the aggregate Sahm rule was triggered, indicating when the (three-month average) national unemployment rate had risen by at least 0.5 percentage points above its low in the previous twelve months. The black line gives the monthly share of states meeting the same criterion at the state level, which we take to indicate that the state is in a recession (defined by the state-level Sahm rule). The chart shows that periods when the national unemployment rate has triggered the Sahm rule corresponds closely to when most states have, but that there are also smaller spikes in the state measure during periods when the national measure is not triggered.
Note: Shaded areas denote months when aggregate Sahm rule was triggered. Aggregate Sahm rule calculated by HAVER and state-level Sahm share calculated from BLS’s Local Area Unemployment Statistics and Current Population Survey. State-level Sahm rule shares are unweighted.
Source: BLS data via Haver Analytics, Authors’ calculations.
Figure 4. Sahm-rule share and population growth
This is a scatter plot showing the relationship between average population growth (x-axis) and the share of months the Sahm rule is triggered (y-axis) at the state level from 1979 to 2019. There is one dot per state (plus Washington, DC). The x-axis displays average annual population growth (percent) from 1979 to 2019, ranging from approximately 0 to 4 percent. The y-axis shows the share of months (percent) that the Sahm rule was triggered, ranging from approximately 0.35 to 0.8. The horizontal dashed line gives the aggregate share of months that the Sahm rule is triggered (about 0.22), and the vertical dashed line gives the aggregate annual population growth average (about 1.15%). The points are fairly dispersed, showing a fairly weak correlation between the two variables.
Note: Sample includes 1979 through 2019. Dashed lines represent aggregate population growth (vertical) and aggregate share of months that Sahm rule is triggered (horizontal).
Source: BLS data via Haver Analytics, Authors’ calculations.
Figure 5: Cross-section Comparison of Population Growth and Selected Macro Outcomes
Left Panel: Population Growth and Unemployment Volatility
| Country | Population growth (Annual average, 2000-2019) | Volatility of the unemployment rate, 2000-2019 |
|---|---|---|
| AU | 1.635168438 | 0.036461942 |
| AS | 0.624536559 | 0.060538202 |
| BE | 0.557049211 | 0.055937071 |
| CA | 1.26787971 | 0.032598743 |
| CL | 1.760158514 | 0.047022383 |
| CZ | 0.254178307 | 0.05674378 |
| DK | 0.58445124 | 0.063835125 |
| ES | -0.129099727 | 0.10814573 |
| FI | 0.474864194 | 0.039619557 |
| HU | -0.073458713 | 0.045202021 |
| IR | 1.452236125 | 0.068846765 |
| IT | 0.327974013 | 0.032659163 |
| JP | 0.136136764 | 0.033869366 |
| KR | 1.078572225 | 0.050006818 |
| NL | 0.62302819 | 0.052013801 |
| NZ | 1.576247839 | 0.058971835 |
| NO | 1.646546105 | 0.059328422 |
| PO | 0.083642162 | 0.052434455 |
| PT | 0.177023817 | 0.043573138 |
| SK | 0.348444766 | 0.042164769 |
| SI | 0.314367196 | 0.061311465 |
| SP | 0.836231897 | 0.04684018 |
| UK | 0.745764816 | 0.033660444 |
| US | 1.035888183 | 0.048880089 |
| GE | 0.06313531 | 0.021183253 |
| FR | 0.606770669 | 0.027695028 |
Right Panel: Population Growth and Share of Quarters with Negative GDP Growth
| Country | Population growth (Annual average, 2000-2019) | Share of quarters with negative GDP growth, 2000-2019 |
|---|---|---|
| AU | 1.635168438 | 3.75 |
| AS | 0.624536559 | 20 |
| BE | 0.557049211 | 12.5 |
| CA | 1.26787971 | 10 |
| CL | 1.760158514 | 17.5 |
| CZ | 0.254178307 | 15 |
| DK | 0.58445124 | 27.5 |
| ES | -0.129099727 | 8.75 |
| FI | 0.474864194 | 30 |
| HU | -0.073458713 | 16.25 |
| IR | 1.452236125 | 33.75 |
| IT | 0.327974013 | 37.5 |
| JP | 0.136136764 | 33.75 |
| KR | 1.078572225 | 5 |
| NL | 0.62302819 | 18.75 |
| NZ | 1.576247839 | 13.75 |
| NO | 1.646546105 | 23.75 |
| PO | 0.083642162 | 13.75 |
| PT | 0.177023817 | 28.75 |
| SK | 0.348444766 | 3.75 |
| SI | 0.314367196 | 6.25 |
| SP | 0.836231897 | 21.25 |
| UK | 0.745764816 | 10 |
| US | 1.035888183 | 12.5 |
| GE | 0.06313531 | 20 |
| FR | 0.606770669 | 19.11764706 |
Note: Population growth is the annualized population growth rate between 2000 and 2019, in percent. Data for France and Germany are from 2003 and 2005, respectively. Volatility of the unemployment rate is the standard deviation of log differences. Share of quarters with negative GDP growth is the ratio of the number of quarters with quarter-on-quarter GDP contraction to the total number of quarters for each country. AS is Austria, ES is Estonia, GE is Germany, SP is Spain, IC is Iceland, UK is the United Kingdom, and labels for other countries use the alpha-2 standard country codes.
Source: OECD via Haver Analytics; Authors’ calculations.
Figure 6: Comparing Low Population Growth Episodes vs. Historical Averages
Chart 1: Unemployment Rate, Mean
| Country | Non low population growth episodes | Low population growth episodes |
|---|---|---|
| BE | 7.523504274 | 7.272792023 |
| CZ | 4.22014652 | 5.60915357 |
| ES | 6.934188034 | 8.356657399 |
| FI | 7.337728938 | 8.14066365 |
| GE | 5.04957265 | 4.926681531 |
| GR | 9.174358974 | 15.37294311 |
| HU | 6.459615385 | 6.483916084 |
| IC | 8.108333333 | 4.676022176 |
| IT | 9.87972028 | 9.198184961 |
| JP | 3.505726496 | 2.649599029 |
| LA | 6.779487179 | 10.22869941 |
| PO | 5.807179487 | 9.506542882 |
| PT | 10.19145299 | 9.43618821 |
| SK | 9.212564103 | 12.17051282 |
| SI | 8.214285714 | 6.24085278 |
| SP | 22.34188034 | 15.28232699 |
Chart 2: Unemployment Rate, Volatility
| Country | Non low population growth episodes | Low population growth episodes |
|---|---|---|
| BE | 0.049101617 | 0.059550664 |
| CZ | 0.057315811 | 0.056377944 |
| ES | 0.078309658 | 0.105728158 |
| FI | 0.06256742 | 0.040943215 |
| GE | 0.030895526 | 0.026092071 |
| GR | 0.05396395 | 0.037876754 |
| HU | 0.025309094 | 0.046446992 |
| IC | 0.072623497 | 0.059797518 |
| IT | 0.027224479 | 0.032158695 |
| JP | 0.023454342 | 0.044427306 |
| LA | 0.047516734 | 0.072678118 |
| PO | 0.047388447 | 0.049452125 |
| PT | 0.02584649 | 0.043050828 |
| SK | 0.01801937 | 0.035439658 |
| SI | 0.05076054 | 0.063296418 |
| SP | 0.012950327 | 0.033794283 |
Note: “Low population growth episodes” are periods in which 3-year population growth was below 0.1 percent, in absolute terms. “Non-Low population growth episodes” are periods in which 3-year population growth was equal or above 0.1 percent. Volatility of the unemployment rate is the standard deviation of log differences. The dotted line denotes the 45-degree line. AS is Austria, ES is Estonia, GE is Germany, SP is Spain, IC is Iceland, UK is the United Kingdom, and labels for other countries use the alpha-2 standard country codes.
Source: OECD via Haver Analytics; Authors’ calculations.