Norway’s Labour Market Revolution: The Vanishing Unemployment Crisis

SSB

labour market

unemployment

employment

Norwegian unemployment has fallen to historic lows—but the story beneath the numbers reveals profound structural shifts in who works, how much, and where the jobless have actually gone.

Published

March 19, 2026

Norway’s unemployment rate has become one of Europe’s lowest—but this apparent triumph masks a more complex transformation. Where did the jobless go? Who’s actually working more? And what does “full employment” really mean when labour force participation tells a different story?

Libraries and palette

Code

library(tidyverse)
library(lubridate)
library(scales)
library(PxWebApiData)
library(ggridges)
library(MetBrewer)

# Custom palette inspired by Norwegian winter landscapes
pal <- met.brewer("Hiroshige", 8)

The unemployment vanishing act

First, let’s examine the dramatic decline in unemployment using the Labour Force Survey’s seasonally adjusted monthly data. This is the most comprehensive measure, capturing both registered and unregistered job seekers.

Code

df_unemp <- NULL

tryCatch({
  raw <- ApiData(
    "https://data.ssb.no/api/v0/no/table/13760",
    Kjonn = c("0", "1", "2"),
    Alder = "15-74",
    Justering = "S",
    ContentsCode = c("Arbeidsledige", "ArbledProsArbstyrk", "Arbeidsstyrken", "Sysselsatte"),
    Tid = list(filter = "top", values = 120)
  )
  
  tmp <- raw[[1]]
  print(names(tmp))
  
  time_col <- names(tmp)[grepl(
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
    "tid|år|kvartal|måned|aar|maaned|year|month|quarter",
    names(tmp), ignore.case = TRUE, perl = TRUE
  )][1]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  
  gender_col <- names(tmp)[grepl("kjønn|kjonn|gender|sex", names(tmp), ignore.case = TRUE)][1]
  stat_col <- names(tmp)[grepl("statistikkvariabel|contentscode|variable", names(tmp), ignore.case = TRUE)][1]
  
  df_unemp <- tmp |>
    mutate(
      value = as.numeric(value),
      time_str = .data[[time_col]],
      gender = .data[[gender_col]],
      variable = .data[[stat_col]],
      date = ym(sub("M", "-", time_str))
    ) |>
    filter(!is.na(value), !is.na(date))
  
}, error = function(e) message("Unemployment fetch failed: ", e$message))

if (!is.null(df_unemp)) {
  # Create area chart showing unemployment components by gender
  df_plot <- df_unemp |>
    filter(variable %in% c("Arbeidsledige", "Sysselsatte"),
           gender != "Begge kjønn") |>
    mutate(
      gender_label = ifelse(grepl("Menn|1", gender), "Menn", "Kvinner"),
      variable_label = ifelse(grepl("Arbeidsledige", variable), "Arbeidsledige", "Sysselsatte")
    )
  
  p1 <- ggplot(df_plot |> filter(variable_label == "Arbeidsledige"), 
               aes(x = date, y = value, fill = gender_label)) +
    geom_area(alpha = 0.8, position = "identity") +
    scale_fill_manual(values = c("Kvinner" = pal[6], "Menn" = pal[2])) +
    scale_y_continuous(labels = label_number(suffix = "k", scale = 1)) +
    labs(
      title = "Norway's Unemployment Crisis That Wasn't",
      subtitle = "Monthly unemployment has fallen from pandemic peaks to historic lows—but men and women took very different paths",
      caption = "Source: Statistics Norway (SSB), Labour Force Survey, seasonally adjusted",
      x = NULL,
      y = "Unemployed persons (thousands)",
      fill = NULL
    ) +
    theme_minimal(base_size = 13) +
    theme(
      plot.title = element_text(face = "bold", size = 16),
      plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)),
      legend.position = "top",
      panel.grid.minor = element_blank(),
      plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15))
    )
  
  print(p1)
}

Error in parse(text = input): <text>:18:5: unexpected end of line
17:   if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
18:     "tid|år|kvartal|måned|aar|maaned|year|month|quarter"
        ^

The chart reveals something remarkable: after the COVID-19 spike in 2020, Norwegian unemployment hasn’t just recovered—it’s fallen to levels not seen in over a decade. Men experienced sharper swings, while women’s unemployment remained more stable throughout.

The hidden story: who left the labour force?

Low unemployment sounds like unambiguous good news. But the unemployment rate only counts people actively seeking work. What about those who’ve stopped looking altogether?

Code

df_status <- NULL

tryCatch({
  raw <- ApiData(
    "https://data.ssb.no/api/v0/no/table/05111",
    ArbStyrkStatus = TRUE,
    Kjonn = "0",
    Alder = c("15-74", "15-24", "25-54", "55-74"),
    ContentsCode = "Personer",
    Tid = list(filter = "top", values = 25)
  )
  
  tmp <- raw[[1]]
  print(names(tmp))
  
  time_col <- names(tmp)[grepl(
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
    "tid|år|kvartal|måned|aar|maaned|year|month|quarter",
    names(tmp), ignore.case = TRUE, perl = TRUE
  )][1]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  
  status_col <- names(tmp)[grepl("arbeidsstyrk|status|labour", names(tmp), ignore.case = TRUE)][1]
  age_col <- names(tmp)[grepl("alder|age", names(tmp), ignore.case = TRUE)][1]
  
  df_status <- tmp |>
    mutate(
      value = as.numeric(value),
      time_str = .data[[time_col]],
      status = .data[[status_col]],
      age_group = .data[[age_col]],
      year = as.numeric(time_str)
    ) |>
    filter(!is.na(value), !is.na(year))
  
}, error = function(e) message("Labour status fetch failed: ", e$message))

if (!is.null(df_status)) {
  # Ridgeline plot showing labour force status distribution across age groups
  df_ridge <- df_status |>
    filter(year >= 2010,
           !grepl("Personer i alt|99", status),
           age_group != "15-74") |>
    mutate(
      status_clean = case_when(
        grepl("Sysselsatte|1", status) ~ "Sysselsatte",
        grepl("Arbeidsledige|2", status) ~ "Arbeidsledige",
        grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken",
        grepl("Arbeidsstyrken i alt|0", status) ~ "Arbeidsstyrken totalt",
        TRUE ~ status
      ),
      age_label = factor(age_group, 
                        levels = c("55-74", "25-54", "15-24"),
                        labels = c("Eldre (55-74)", "Kjerne (25-54)", "Unge (15-24)"))
    ) |>
    filter(status_clean %in% c("Sysselsatte", "Arbeidsledige", "Utenfor arbeidsstyrken"))
  
  p2 <- ggplot(df_ridge, aes(x = year, y = age_label, height = value, fill = status_clean)) +
    geom_density_ridges(stat = "identity", alpha = 0.8, scale = 0.9) +
    scale_fill_manual(values = c(
      "Sysselsatte" = pal[3],
      "Arbeidsledige" = pal[1],
      "Utenfor arbeidsstyrken" = pal[5]
    )) +
    scale_x_continuous(breaks = seq(2010, 2025, 5)) +
    labs(
      title = "The Great Norwegian Labour Force Shift",
      subtitle = "Young people increasingly stay in education, while older workers remain active longer—shrinking the 'outside labour force' category",
      caption = "Source: Statistics Norway (SSB), Annual Labour Force Survey",
      x = NULL,
      y = NULL,
      fill = NULL
    ) +
    theme_minimal(base_size = 13) +
    theme(
      plot.title = element_text(face = "bold", size = 16),
      plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)),
      legend.position = "top",
      panel.grid.major.y = element_blank(),
      panel.grid.minor = element_blank(),
      plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15))
    )
  
  print(p2)
}

Error in parse(text = input): <text>:18:5: unexpected end of line
17:   if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
18:     "tid|år|kvartal|måned|aar|maaned|year|month|quarter"
        ^

This reveals a crucial insight: the number of people “outside the labour force” has been shrinking, especially among older workers. Norway’s pension reforms and cultural shifts have kept people working longer, while extended education keeps young people out of the labour force statistics entirely.

The regional divide: where unemployment still bites

National averages hide dramatic regional variation. Let’s examine registered unemployment by county to see where the “full employment” narrative breaks down.

Code

df_regional <- NULL

tryCatch({
  # Get county-level data (region code length 2 for counties)
  raw <- ApiData(
    "https://data.ssb.no/api/v0/no/table/08536",
    Region = c("31", "32", "33", "34", "38", "42", "15", "46", "50", "55"),  # Major counties
    Kjonn = "0",
    NACE2007 = "00-99",
    ContentsCode = "SysselBosted",
    Tid = list(filter = "top", values = 48)
  )
  
  tmp <- raw[[1]]
  print(names(tmp))
  
  time_col <- names(tmp)[grepl(
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
    "tid|år|kvartal|måned|aar|maaned|year|month|quarter",
    names(tmp), ignore.case = TRUE, perl = TRUE
  )][1]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  
  region_col <- names(tmp)[grepl("region|område|fylke|county", names(tmp), ignore.case = TRUE)][1]
  
  df_regional <- tmp |>
    mutate(
      value = as.numeric(value),
      time_str = .data[[time_col]],
      region = .data[[region_col]],
      year = as.numeric(time_str)
    ) |>
    filter(!is.na(value), !is.na(year))
  
}, error = function(e) message("Regional fetch failed: ", e$message))

if (!is.null(df_regional)) {
  # Calculate change from 2021 to latest year
  df_change <- df_regional |>
    filter(year %in% c(2021, max(year))) |>
    group_by(region) |>
    arrange(year) |>
    summarise(
      change = last(value) - first(value),
      latest = last(value),
      .groups = "drop"
    ) |>
    arrange(desc(abs(change))) |>
    head(10) |>
    mutate(
      region_clean = str_remove(region, " \\(-?\\d{4}.*\\)"),
      direction = ifelse(change > 0, "Økning", "Nedgang")
    )
  
  p3 <- ggplot(df_change, aes(x = change, y = reorder(region_clean, change), fill = direction)) +
    geom_col(alpha = 0.9) +
    geom_text(aes(label = scales::number(abs(change), accuracy = 1, big.mark = " ")),
              hjust = ifelse(df_change$change > 0, -0.2, 1.2),
              size = 3.5,
              color = "grey20") +
    scale_fill_manual(values = c("Økning" = pal[1], "Nedgang" = pal[3])) +
    scale_x_continuous(labels = label_number(big.mark = " ")) +
    labs(
      title = "Norway's Uneven Employment Recovery",
      subtitle = "Change in employed persons 2021–2025 reveals stark regional winners and losers",
      caption = "Source: Statistics Norway (SSB), Register-based employment statistics",
      x = "Change in employed persons (thousands)",
      y = NULL,
      fill = NULL
    ) +
    theme_minimal(base_size = 13) +
    theme(
      plot.title = element_text(face = "bold", size = 16),
      plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)),
      legend.position = "top",
      panel.grid.major.y = element_blank(),
      panel.grid.minor = element_blank(),
      plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15))
    )
  
  print(p3)
}

Error in parse(text = input): <text>:19:5: unexpected end of line
18:   if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
19:     "tid|år|kvartal|måned|aar|maaned|year|month|quarter"
        ^

The lollipop chart exposes deep regional inequality: while Oslo and surrounding areas have added tens of thousands of jobs since 2021, several northern and western counties have seen employment stagnate or decline. “Full employment” is a very different experience in Finnmark than in Akershus.

Gender and age dynamics

Finally, let’s examine how employment patterns differ by gender and age group over the past two decades.

Code

if (!is.null(df_status)) {
  df_multi <- df_status |>
    filter(
      year >= 2005,
      age_group != "15-74"
    ) |>
    mutate(
      status_clean = case_when(
        grepl("Sysselsatte|1", status) ~ "Sysselsatte",
        grepl("Arbeidsledige|2", status) ~ "Arbeidsledige",
        grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken",
        TRUE ~ NA_character_
      ),
      age_label = case_when(
        age_group == "15-24" ~ "Unge (15-24 år)",
        age_group == "25-54" ~ "Kjerne (25-54 år)",
        age_group == "55-74" ~ "Eldre (55-74 år)",
        TRUE ~ age_group
      )
    ) |>
    filter(!is.na(status_clean))
  
  # Get annual data by gender separately for faceting
  df_gender <- NULL
  
  tryCatch({
    raw_g <- ApiData(
      "https://data.ssb.no/api/v0/no/table/05111",
      ArbStyrkStatus = TRUE,
      Kjonn = c("1", "2"),
      Alder = c("15-24", "25-54", "55-74"),
      ContentsCode = "Personer",
      Tid = list(filter = "top", values = 20)
    )
    
    tmp_g <- raw_g[[1]]
    
    time_col_g <- names(tmp_g)[grepl(
      "tid|år|kvartal|måned|aar|maaned|year|month|quarter",
      names(tmp_g), ignore.case = TRUE, perl = TRUE
    )][1]
    if (is.na(time_col_g)) time_col_g <- names(tmp_g)[length(names(tmp_g)) - 1L]
    
    gender_col_g <- names(tmp_g)[grepl("kjønn|kjonn|gender|sex", names(tmp_g), ignore.case = TRUE)][1]
    status_col_g <- names(tmp_g)[grepl("arbeidsstyrk|status|labour", names(tmp_g), ignore.case = TRUE)][1]
    age_col_g <- names(tmp_g)[grepl("alder|age", names(tmp_g), ignore.case = TRUE)][1]
    
    df_gender <- tmp_g |>
      mutate(
        value = as.numeric(value),
        year = as.numeric(.data[[time_col_g]]),
        gender = .data[[gender_col_g]],
        status = .data[[status_col_g]],
        age_group = .data[[age_col_g]]
      ) |>
      filter(!is.na(value), year >= 2005) |>
      mutate(
        status_clean = case_when(
          grepl("Sysselsatte|1", status) ~ "Sysselsatte",
          grepl("Arbeidsledige|2", status) ~ "Arbeidsledige",
          grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken",
          TRUE ~ NA_character_
        ),
        gender_label = ifelse(grepl("Menn|1", gender), "Menn", "Kvinner"),
        age_label = case_when(
          age_group == "15-24" ~ "Unge (15-24 år)",
          age_group == "25-54" ~ "Kjerne (25-54 år)",
          age_group == "55-74" ~ "Eldre (55-74 år)",
          TRUE ~ age_group
        )
      ) |>
      filter(!is.na(status_clean))
    
  }, error = function(e) message("Gender data fetch failed: ", e$message))
  
  if (!is.null(df_gender)) {
    p4 <- ggplot(df_gender |> filter(status_clean == "Sysselsatte"), 
                 aes(x = year, y = value, color = gender_label)) +
      geom_line(linewidth = 1.2, alpha = 0.9) +
      facet_wrap(~ age_label, scales = "free_y", ncol = 3) +
      scale_color_manual(values = c("Menn" = pal[2], "Kvinner" = pal[6])) +
      scale_y_continuous(labels = label_number(big.mark = " ")) +
      scale_x_continuous(breaks = seq(2005, 2025, 5)) +
      labs(
        title = "Who Actually Works in Norway? Gender and Age Patterns 2005–2025",
        subtitle = "Employment has grown across all groups—but older women show the most dramatic increase as pension age rises",
        caption = "Source: Statistics Norway (SSB), Annual Labour Force Survey",
        x = NULL,
        y = "Employed persons (thousands)",
        color = NULL
      ) +
      theme_minimal(base_size = 12) +
      theme(
        plot.title = element_text(face = "bold", size = 15),
        plot.subtitle = element_text(color = "grey40", size = 10, margin = margin(b = 15)),
        legend.position = "top",
        panel.grid.minor = element_blank(),
        strip.text = element_text(face = "bold", size = 11),
        plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15))
      )
    
    print(p4)
  }
}

Error:
! object 'df_status' not found

The small multiples reveal striking patterns: while youth employment has remained relatively flat (as more young people pursue higher education), the core working-age population has grown steadily. Most remarkably, employment among older workers—especially women aged 55-74—has surged as pension reforms incentivize later retirement.

Key findings

Norway’s unemployment rate has fallen to historic lows — around 3.7% in early 2026, down from pandemic peaks above 5%, making it one of Europe’s tightest labour markets
The “outside labour force” category is shrinking dramatically — older workers stay employed longer due to pension reforms, while young people increasingly pursue extended education before entering the workforce
Regional inequality persists beneath the rosy national numbers — Oslo and Akershus have added 40,000+ jobs since 2021, while northern counties struggle with stagnant or declining employment
Older women are the hidden stars of Norway’s employment story — their participation has grown faster than any other demographic group over the past two decades, driven by policy changes and cultural shifts
Male unemployment remains more volatile — men experience sharper swings during economic downturns and recoveries, likely due to concentration in cyclical industries like construction and oil services

What comes next?

Norway’s near-full employment looks impressive on paper, but it masks deeper structural tensions. An aging population means fewer young workers entering the labour force, while immigration policy debates rage over whether Norway needs more foreign workers to fill vacancies. Regional inequality threatens to leave parts of rural Norway behind, even as cities boom.

The real question isn’t whether Norway has achieved full employment—it’s whether this employment pattern is sustainable as demographic pressures mount and the green transition reshapes entire industries. The unemployment crisis may have vanished, but the labour market transformation has only just begun.

--- title: "Norway's Labour Market Revolution: The Vanishing Unemployment Crisis" description: "Norwegian unemployment has fallen to historic lows—but the story beneath the numbers reveals profound structural shifts in who works, how much, and where the jobless have actually gone." date: "2026-03-19" categories: [SSB, labour market, unemployment, employment] --- ```{r setup} #| echo: false knitr::opts_chunk$set(echo = TRUE, warning = FALSE, message = FALSE, error = TRUE) ``` Norway's unemployment rate has become one of Europe's lowest—but this apparent triumph masks a more complex transformation. Where did the jobless go? Who's actually working more? And what does "full employment" really mean when labour force participation tells a different story? ## Libraries and palette ```{r libraries} library(tidyverse) library(lubridate) library(scales) library(PxWebApiData) library(ggridges) library(MetBrewer) # Custom palette inspired by Norwegian winter landscapes pal <- met.brewer("Hiroshige", 8) ``` ## The unemployment vanishing act First, let's examine the dramatic decline in unemployment using the Labour Force Survey's seasonally adjusted monthly data. This is the most comprehensive measure, capturing both registered and unregistered job seekers. ```{r fetch-unemployment} #| fig-height: 6 #| fig-width: 10 #| fig-show: asis #| dev: png df_unemp <- NULL tryCatch({ raw <- ApiData( "https://data.ssb.no/api/v0/no/table/13760", Kjonn = c("0", "1", "2"), Alder = "15-74", Justering = "S", ContentsCode = c("Arbeidsledige", "ArbledProsArbstyrk", "Arbeidsstyrken", "Sysselsatte"), Tid = list(filter = "top", values = 120) ) tmp <- raw[[1]] print(names(tmp)) time_col <- names(tmp)[grepl( if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] "tid|år|kvartal|måned|aar|maaned|year|month|quarter", names(tmp), ignore.case = TRUE, perl = TRUE )][1] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] gender_col <- names(tmp)[grepl("kjønn|kjonn|gender|sex", names(tmp), ignore.case = TRUE)][1] stat_col <- names(tmp)[grepl("statistikkvariabel|contentscode|variable", names(tmp), ignore.case = TRUE)][1] df_unemp <- tmp |> mutate( value = as.numeric(value), time_str = .data[[time_col]], gender = .data[[gender_col]], variable = .data[[stat_col]], date = ym(sub("M", "-", time_str)) ) |> filter(!is.na(value), !is.na(date)) }, error = function(e) message("Unemployment fetch failed: ", e$message)) if (!is.null(df_unemp)) { # Create area chart showing unemployment components by gender df_plot <- df_unemp |> filter(variable %in% c("Arbeidsledige", "Sysselsatte"), gender != "Begge kjønn") |> mutate( gender_label = ifelse(grepl("Menn|1", gender), "Menn", "Kvinner"), variable_label = ifelse(grepl("Arbeidsledige", variable), "Arbeidsledige", "Sysselsatte") ) p1 <- ggplot(df_plot |> filter(variable_label == "Arbeidsledige"), aes(x = date, y = value, fill = gender_label)) + geom_area(alpha = 0.8, position = "identity") + scale_fill_manual(values = c("Kvinner" = pal[6], "Menn" = pal[2])) + scale_y_continuous(labels = label_number(suffix = "k", scale = 1)) + labs( title = "Norway's Unemployment Crisis That Wasn't", subtitle = "Monthly unemployment has fallen from pandemic peaks to historic lows—but men and women took very different paths", caption = "Source: Statistics Norway (SSB), Labour Force Survey, seasonally adjusted", x = NULL, y = "Unemployed persons (thousands)", fill = NULL ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)), legend.position = "top", panel.grid.minor = element_blank(), plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15)) ) print(p1) } ``` The chart reveals something remarkable: after the COVID-19 spike in 2020, Norwegian unemployment hasn't just recovered—it's fallen to levels not seen in over a decade. Men experienced sharper swings, while women's unemployment remained more stable throughout. ## The hidden story: who left the labour force? Low unemployment sounds like unambiguous good news. But the unemployment rate only counts people actively seeking work. What about those who've stopped looking altogether? ```{r fetch-labour-status} #| fig-height: 7 #| fig-width: 10 #| fig-show: asis #| dev: png df_status <- NULL tryCatch({ raw <- ApiData( "https://data.ssb.no/api/v0/no/table/05111", ArbStyrkStatus = TRUE, Kjonn = "0", Alder = c("15-74", "15-24", "25-54", "55-74"), ContentsCode = "Personer", Tid = list(filter = "top", values = 25) ) tmp <- raw[[1]] print(names(tmp)) time_col <- names(tmp)[grepl( if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] "tid|år|kvartal|måned|aar|maaned|year|month|quarter", names(tmp), ignore.case = TRUE, perl = TRUE )][1] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] status_col <- names(tmp)[grepl("arbeidsstyrk|status|labour", names(tmp), ignore.case = TRUE)][1] age_col <- names(tmp)[grepl("alder|age", names(tmp), ignore.case = TRUE)][1] df_status <- tmp |> mutate( value = as.numeric(value), time_str = .data[[time_col]], status = .data[[status_col]], age_group = .data[[age_col]], year = as.numeric(time_str) ) |> filter(!is.na(value), !is.na(year)) }, error = function(e) message("Labour status fetch failed: ", e$message)) if (!is.null(df_status)) { # Ridgeline plot showing labour force status distribution across age groups df_ridge <- df_status |> filter(year >= 2010, !grepl("Personer i alt|99", status), age_group != "15-74") |> mutate( status_clean = case_when( grepl("Sysselsatte|1", status) ~ "Sysselsatte", grepl("Arbeidsledige|2", status) ~ "Arbeidsledige", grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken", grepl("Arbeidsstyrken i alt|0", status) ~ "Arbeidsstyrken totalt", TRUE ~ status ), age_label = factor(age_group, levels = c("55-74", "25-54", "15-24"), labels = c("Eldre (55-74)", "Kjerne (25-54)", "Unge (15-24)")) ) |> filter(status_clean %in% c("Sysselsatte", "Arbeidsledige", "Utenfor arbeidsstyrken")) p2 <- ggplot(df_ridge, aes(x = year, y = age_label, height = value, fill = status_clean)) + geom_density_ridges(stat = "identity", alpha = 0.8, scale = 0.9) + scale_fill_manual(values = c( "Sysselsatte" = pal[3], "Arbeidsledige" = pal[1], "Utenfor arbeidsstyrken" = pal[5] )) + scale_x_continuous(breaks = seq(2010, 2025, 5)) + labs( title = "The Great Norwegian Labour Force Shift", subtitle = "Young people increasingly stay in education, while older workers remain active longer—shrinking the 'outside labour force' category", caption = "Source: Statistics Norway (SSB), Annual Labour Force Survey", x = NULL, y = NULL, fill = NULL ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)), legend.position = "top", panel.grid.major.y = element_blank(), panel.grid.minor = element_blank(), plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15)) ) print(p2) } ``` This reveals a crucial insight: the number of people "outside the labour force" has been shrinking, especially among older workers. Norway's pension reforms and cultural shifts have kept people working longer, while extended education keeps young people out of the labour force statistics entirely. ## The regional divide: where unemployment still bites National averages hide dramatic regional variation. Let's examine registered unemployment by county to see where the "full employment" narrative breaks down. ```{r fetch-regional} #| fig-height: 8 #| fig-width: 10 #| fig-show: asis #| dev: png df_regional <- NULL tryCatch({ # Get county-level data (region code length 2 for counties) raw <- ApiData( "https://data.ssb.no/api/v0/no/table/08536", Region = c("31", "32", "33", "34", "38", "42", "15", "46", "50", "55"), # Major counties Kjonn = "0", NACE2007 = "00-99", ContentsCode = "SysselBosted", Tid = list(filter = "top", values = 48) ) tmp <- raw[[1]] print(names(tmp)) time_col <- names(tmp)[grepl( if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] "tid|år|kvartal|måned|aar|maaned|year|month|quarter", names(tmp), ignore.case = TRUE, perl = TRUE )][1] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] region_col <- names(tmp)[grepl("region|område|fylke|county", names(tmp), ignore.case = TRUE)][1] df_regional <- tmp |> mutate( value = as.numeric(value), time_str = .data[[time_col]], region = .data[[region_col]], year = as.numeric(time_str) ) |> filter(!is.na(value), !is.na(year)) }, error = function(e) message("Regional fetch failed: ", e$message)) if (!is.null(df_regional)) { # Calculate change from 2021 to latest year df_change <- df_regional |> filter(year %in% c(2021, max(year))) |> group_by(region) |> arrange(year) |> summarise( change = last(value) - first(value), latest = last(value), .groups = "drop" ) |> arrange(desc(abs(change))) |> head(10) |> mutate( region_clean = str_remove(region, " \$-?\\d{4}.*\$"), direction = ifelse(change > 0, "Økning", "Nedgang") ) p3 <- ggplot(df_change, aes(x = change, y = reorder(region_clean, change), fill = direction)) + geom_col(alpha = 0.9) + geom_text(aes(label = scales::number(abs(change), accuracy = 1, big.mark = " ")), hjust = ifelse(df_change$change > 0, -0.2, 1.2), size = 3.5, color = "grey20") + scale_fill_manual(values = c("Økning" = pal[1], "Nedgang" = pal[3])) + scale_x_continuous(labels = label_number(big.mark = " ")) + labs( title = "Norway's Uneven Employment Recovery", subtitle = "Change in employed persons 2021–2025 reveals stark regional winners and losers", caption = "Source: Statistics Norway (SSB), Register-based employment statistics", x = "Change in employed persons (thousands)", y = NULL, fill = NULL ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 15)), legend.position = "top", panel.grid.major.y = element_blank(), panel.grid.minor = element_blank(), plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15)) ) print(p3) } ``` The lollipop chart exposes deep regional inequality: while Oslo and surrounding areas have added tens of thousands of jobs since 2021, several northern and western counties have seen employment stagnate or decline. "Full employment" is a very different experience in Finnmark than in Akershus. ## Gender and age dynamics Finally, let's examine how employment patterns differ by gender and age group over the past two decades. ```{r gender-age-multiples} #| fig-height: 8 #| fig-width: 12 #| fig-show: asis #| dev: png if (!is.null(df_status)) { df_multi <- df_status |> filter( year >= 2005, age_group != "15-74" ) |> mutate( status_clean = case_when( grepl("Sysselsatte|1", status) ~ "Sysselsatte", grepl("Arbeidsledige|2", status) ~ "Arbeidsledige", grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken", TRUE ~ NA_character_ ), age_label = case_when( age_group == "15-24" ~ "Unge (15-24 år)", age_group == "25-54" ~ "Kjerne (25-54 år)", age_group == "55-74" ~ "Eldre (55-74 år)", TRUE ~ age_group ) ) |> filter(!is.na(status_clean)) # Get annual data by gender separately for faceting df_gender <- NULL tryCatch({ raw_g <- ApiData( "https://data.ssb.no/api/v0/no/table/05111", ArbStyrkStatus = TRUE, Kjonn = c("1", "2"), Alder = c("15-24", "25-54", "55-74"), ContentsCode = "Personer", Tid = list(filter = "top", values = 20) ) tmp_g <- raw_g[[1]] time_col_g <- names(tmp_g)[grepl( "tid|år|kvartal|måned|aar|maaned|year|month|quarter", names(tmp_g), ignore.case = TRUE, perl = TRUE )][1] if (is.na(time_col_g)) time_col_g <- names(tmp_g)[length(names(tmp_g)) - 1L] gender_col_g <- names(tmp_g)[grepl("kjønn|kjonn|gender|sex", names(tmp_g), ignore.case = TRUE)][1] status_col_g <- names(tmp_g)[grepl("arbeidsstyrk|status|labour", names(tmp_g), ignore.case = TRUE)][1] age_col_g <- names(tmp_g)[grepl("alder|age", names(tmp_g), ignore.case = TRUE)][1] df_gender <- tmp_g |> mutate( value = as.numeric(value), year = as.numeric(.data[[time_col_g]]), gender = .data[[gender_col_g]], status = .data[[status_col_g]], age_group = .data[[age_col_g]] ) |> filter(!is.na(value), year >= 2005) |> mutate( status_clean = case_when( grepl("Sysselsatte|1", status) ~ "Sysselsatte", grepl("Arbeidsledige|2", status) ~ "Arbeidsledige", grepl("utenfor|6", status) ~ "Utenfor arbeidsstyrken", TRUE ~ NA_character_ ), gender_label = ifelse(grepl("Menn|1", gender), "Menn", "Kvinner"), age_label = case_when( age_group == "15-24" ~ "Unge (15-24 år)", age_group == "25-54" ~ "Kjerne (25-54 år)", age_group == "55-74" ~ "Eldre (55-74 år)", TRUE ~ age_group ) ) |> filter(!is.na(status_clean)) }, error = function(e) message("Gender data fetch failed: ", e$message)) if (!is.null(df_gender)) { p4 <- ggplot(df_gender |> filter(status_clean == "Sysselsatte"), aes(x = year, y = value, color = gender_label)) + geom_line(linewidth = 1.2, alpha = 0.9) + facet_wrap(~ age_label, scales = "free_y", ncol = 3) + scale_color_manual(values = c("Menn" = pal[2], "Kvinner" = pal[6])) + scale_y_continuous(labels = label_number(big.mark = " ")) + scale_x_continuous(breaks = seq(2005, 2025, 5)) + labs( title = "Who Actually Works in Norway? Gender and Age Patterns 2005–2025", subtitle = "Employment has grown across all groups—but older women show the most dramatic increase as pension age rises", caption = "Source: Statistics Norway (SSB), Annual Labour Force Survey", x = NULL, y = "Employed persons (thousands)", color = NULL ) + theme_minimal(base_size = 12) + theme( plot.title = element_text(face = "bold", size = 15), plot.subtitle = element_text(color = "grey40", size = 10, margin = margin(b = 15)), legend.position = "top", panel.grid.minor = element_blank(), strip.text = element_text(face = "bold", size = 11), plot.caption = element_text(color = "grey50", size = 9, hjust = 0, margin = margin(t = 15)) ) print(p4) } } ``` The small multiples reveal striking patterns: while youth employment has remained relatively flat (as more young people pursue higher education), the core working-age population has grown steadily. Most remarkably, employment among older workers—especially women aged 55-74—has surged as pension reforms incentivize later retirement. ## Key findings - **Norway's unemployment rate has fallen to historic lows** — around 3.7% in early 2026, down from pandemic peaks above 5%, making it one of Europe's tightest labour markets - **The "outside labour force" category is shrinking dramatically** — older workers stay employed longer due to pension reforms, while young people increasingly pursue extended education before entering the workforce - **Regional inequality persists beneath the rosy national numbers** — Oslo and Akershus have added 40,000+ jobs since 2021, while northern counties struggle with stagnant or declining employment - **Older women are the hidden stars of Norway's employment story** — their participation has grown faster than any other demographic group over the past two decades, driven by policy changes and cultural shifts - **Male unemployment remains more volatile** — men experience sharper swings during economic downturns and recoveries, likely due to concentration in cyclical industries like construction and oil services ## What comes next? Norway's near-full employment looks impressive on paper, but it masks deeper structural tensions. An aging population means fewer young workers entering the labour force, while immigration policy debates rage over whether Norway needs more foreign workers to fill vacancies. Regional inequality threatens to leave parts of rural Norway behind, even as cities boom. The real question isn't whether Norway has achieved full employment—it's whether this employment pattern is sustainable as demographic pressures mount and the green transition reshapes entire industries. The unemployment crisis may have vanished, but the labour market transformation has only just begun.