South Korea Average Monthly Wage per Regular Worker

The youth unemployment rate, the average monthly wage, and average weekly hours worked are three not-seasonally-adjusted labor-market indicators that KRED carries without transformation, each recorded exactly as published. All three rest on a standard labor-force framework (Hussmanns, Mehran, and Verma 1990), which classifies each person of working age as employed, unemployed, or outside the labor force by activity status, availability, and job-search criteria. The youth unemployment rate divides the unemployed aged fifteen to twenty-nine by the economically active population in that age band, while the wage and hours measures are per-worker averages over the employed.

Each indicator is read as the net outcome of continuous flows of entry, exit, and spell duration (Clark and Summers 1979), so that a single monthly reading summarizes the churn beneath the surface level.

Within the cyclical-indicator tradition these labor readings are comoving series whose turning points cluster around aggregate reference cycles (Burns and Mitchell 1946). They are classified by their timing relative to the cycle (Mitchell and Burns 1938), their turns are dated by a reproducible algorithm (Bry and Boschan 1971), and they are placed within the diffusion- and composite-index apparatus (Moore 1961).

Labor aggregates such as these are recast as observable shadows of a latent state of the economy (Stock and Watson 1989; Stock and Watson 1991). Their measurement and timing properties are catalogued systematically (Zarnowitz 1992), and each value is treated as a monthly current-conditions signal (Shiskin 1961). Such atheoretical measurement, however, carries meaning only against an explicit model of the labor market (Koopmans 1947).

KRED reports each figure exactly as constructed and applies no transformation, neither rescaling, deflating, smoothing, annualizing, nor seasonal adjustment. Because the underlying series are not seasonally adjusted, each retains its within-year seasonal pattern, and the raw passthrough leaves that pattern intact.

The measurement basis is the household and establishment labor survey, in which each respondent's status is translated into a discrete activity category and per-worker earnings and hours are recorded under the operational definitions of the active population, the employed, and the unemployed (Hussmanns, Mehran, and Verma 1990). The flow accounting that links those statuses across months decomposes the change in each figure into entry, exit, and duration (Clark and Summers 1979).

The exercise belongs to the lineage of qualitative tendency surveys. The link between categorical survey replies and official output statistics was first modeled (Anderson 1952), the quantification of up-same-down responses was founded (Theil 1952), a probability method with indifference thresholds was added (Carlson and Parkin 1975), and competing balance and regression schemes were reviewed (Nardo 2003).

The cyclical reading of the resulting figures follows a reproducible turning-point algorithm that dates their specific cycles (Bry and Boschan 1971) and a codified embedding in diffusion and composite indexes (Moore 1961). The single-index dynamic-factor representation is formalized (Stock and Watson 1991), as is the monthly real-time construction (Shiskin 1961). This measurement procedure nonetheless presupposes an economic model for its interpretation (Koopmans 1947).

As measures of labor-market slack, pay, and effort, these indicators anchor the reading of the business cycle. Their clustered turning points define the cycle as comoving series (Burns and Mitchell 1946), and each is classified by its timing relative to a turn (Mitchell and Burns 1938).

A rising youth unemployment rate or a softening in wages and hours signals a deepening of cyclical weakness rather than an idiosyncratic shock, because labor aggregates are treated as noisy readings of the latent state of the economy (Stock and Watson 1989; Stock and Watson 1991).

The indicators play a standard role in composite cyclical assessment. Their place within diffusion- and composite-indicator systems is documented (Moore 1961; Zarnowitz 1992), turning-point dating separates cyclical from incidental movements (Bry and Boschan 1971), and their use as current-conditions gauges is formalized as a monthly update (Shiskin 1961).

An identical headline figure can carry different implications when the underlying flow mix differs. A stable surface value can mask shifting entry, exit, and duration flows that carry distinct welfare meaning (Clark and Summers 1979), and the population over which each measure is defined is itself bounded (Hussmanns, Mehran, and Verma 1990).

Labor-market conditions govern household income expectations, so these readings link to spending behavior (Katona 1951). Labor and confidence readings carry forward-looking content for consumption (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004). The indicators, however, inform but do not identify the mechanism (Koopmans 1947).

For markets these labor readings are a high-frequency read on cyclical timing, and their classification by cycle timing governs how a given print maps into expectations for the policy path (Mitchell and Burns 1938; Burns and Mitchell 1946).

Fixed-income and rate markets track them as components of a latent-state estimate (Stock and Watson 1989; Stock and Watson 1991). How much weight a surprise deserves is disciplined by the series' forecasting record (Zarnowitz 1992).

An unexpected move in the youth unemployment rate, in wages, or in hours shifts the expected trajectory of growth and inflation, because labor and confidence readings carry independent predictive content for household demand (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004). Wage and hours prints in particular carry the labor-cost signal that rate markets read into the inflation path.

Each release is positioned within a composite and real-time signaling frame (Moore 1961; Shiskin 1961). Phase transitions are flagged by turning-point dating (Bry and Boschan 1971), and a stable surface value must not be over-read against the flow decomposition (Clark and Summers 1979).

The labor signal connects to consumer attitudes that feed risk appetite (Katona 1951). The same number nonetheless supports divergent trades depending on the model the trader imposes (Koopmans 1947).

KRWAGE is published as a not-seasonally-adjusted nominal wage level, and the appropriate statistical object is its logarithm. The sample is short at 76 monthly observations spanning 2020-01 to 2026-04, and the unit-root triplet is fitted on the log level with a constant and trend, so the limited power of these tests on so few observations is kept in view.

The augmented Dickey-Fuller regression of Dickey and Fuller (1979), in the lag-augmented form of Said and Dickey (1984), does not reject the unit root with a statistic of −2.381 at p = 0.390, while the nonparametric test of Phillips and Perron (1988) rejects it with a statistic of −8.473 at p = 0.000 and the stationarity-null test fails to reject trend stationarity with a statistic of 0.026 at p > 0.10 (Kwiatkowski et al. 1992), so the three legs disagree and no clean order of integration is assigned. The GLS-detrended escalation of Elliott, Rothenberg, and Stock (1996) with the modified M-tests of Ng and Perron (2001) is reserved for such ambiguous readings and never enters as a fourth vote, so neither overturns the honest non-verdict, and the short span rather than a settled stationarity is the more economical reading of the split.

Because the log level is not a clean stationary object, the portmanteau and the break search are run on its first difference, since a level portmanteau reads mechanical near-unit autocorrelation and a level mean-break search spuriously segments a stochastic trend (Perron 1989; Hamilton 2018). The portmanteau statistic of Ljung and Box (1978), refining the form of Box and Pierce (1970), rejects the white-noise null on the differenced log level, with Q = 53.40 at lag 12 and Q = 106.79 at lag 24, each at p = 0.000, and the multiple-break procedure of Bai and Perron (1998), computed by the dynamic-programming algorithm of Bai and Perron (2003), locates no break in the mean of the differenced series.

Seasonal tests are appropriate on this unadjusted monthly series. The seasonal-unit-root test of Hylleberg et al. (1990), with the monthly mechanics of Beaulieu and Miron (1993), rejects unit roots at the seasonal frequencies with a statistic of 9.073 at p = 0.000, while the seasonal dummies are jointly significant with an F statistic of 17.013 at p = 0.000, so the seasonality is present and deterministic rather than stochastic, consistent with the stationary-seasonality reading of Canova and Hansen (1995) and with the not-seasonally-adjusted character of the series (Ghysels and Osborn 2001).