South Korea Central Government Total Expenditure

The Korean government-finance flows KRGOVREV, KRGOVEXP, and KRTAXREV are raw series recorded as reported without any transformation, the won magnitudes of central-government receipts and outlays as they accumulate through the fiscal year. KRGOVREV is total revenue, KRGOVEXP is total expenditure, and KRTAXREV is national tax revenue, the compulsory part of the current receipts that sit inside total revenue, each expressed in billions of won and compiled as a within-year cumulative flow that resets at the start of each year. The consolidated fiscal balance published as KRFISCAL is the net of total revenue over total expenditure and net lending, so these series are the gross sides of that same balance.

Both sides of the account are national-accounting magnitudes whose classification is fixed by the double-entry construction of the government sector (Meade and Stone 1941; Stone 1947), where public receipts and outlays enter as current and capital transactions of the sector rather than as a cash residual. The aggregate these flows feed is defined within the national-income framework that treats government activity as a distinct account (Kuznets 1941), and the measurement of that activity as an object of estimation in its own right has a long empirical lineage (Fabricant 1940). Read as financial magnitudes, the receipts and outlays of the public sector are entered against the matching claims of the counterparties who pay and receive them (Copeland 1949), and the sectoral position that results is tabulated in the financial account of the integrated social-accounting system (Stone 1966).

A tax receipt is a transfer to the public sector rather than a purchase, so KRTAXREV measures the compulsory part of the receipts side and moves with the base on which it is levied rather than with the volume of public services rendered. Expenditure conversely covers purchases, transfers, and interest, and interest is the price that equates present and future public income (Fisher 1930). Because revenue and outlay are flows while the liability they leave behind is a stock, the three series connect to outstanding government debt through the accounting identity that cumulates deficits (Obstfeld and Rogoff 1995), and the position that identity produces belongs to the family of national balance-sheet measures (Goldsmith, Lipsey, and Mendelson 1963; Lane and Milesi-Ferretti 2001).

None of the three carries a seasonal adjustment, and each is recorded here exactly as compiled, at monthly frequency on its cumulative basis (Dawson 1996). Because the receipts side responds to activity while much of the outlay side is committed in advance, the gap between them widens and narrows with the cycle, a co-movement in the tradition of business-cycle measurement (Burns and Mitchell 1946), and the debt-to-activity ratio that the gap eventually implies is a structural gauge in the long-run record (Schularick and Taylor 2012).

KRED applies no transformation to KRGOVREV, KRGOVEXP, or KRTAXREV, neither deflating, re-scaling, smoothing, annualizing, nor seasonally adjusting them, so each published number is the compiled figure exactly as measured in billions of won. All three are reproduced on their native cumulative year-to-date basis, and none is decomposed into monthly increments or converted into a rolling twelve-month total.

The classification that assigns a receipt to revenue and a payment to expenditure follows the double-entry construction of the government account (Meade and Stone 1941; Stone 1947), within which the tax component is separated from the other current receipts. The account itself is read as the government sector of the national-income framework (Kuznets 1941), whose boundary between the public sector and the rest of the economy rests on the early empirical delineation of public activity (Fabricant 1940). The counterpart entries are those of moneyflows accounting, in which every payment to or from the public sector is recorded simultaneously against the counterparty (Copeland 1949), later implemented as a full empirical system (Copeland 1952).

The sectoring and instrument conventions that keep the flows consistent with the corresponding stocks follow the handbook treatment of financial accounts (Dawson 1996), and the stock side against which the flows are checked is the sectoral balance sheet (Goldsmith, Lipsey, and Mendelson 1963). Within the integrated social-accounting system the three flows are entries of the government sector rather than modelled quantities (Stone 1966), and the reliability of any such estimate is itself treated as a measured property of the accounts (Stone, Champernowne, and Meade 1942).

Because a cumulative series is an accounting sum and not a smoothed one, each observation is the running total of the flows recorded since January, and the twelfth observation of a year is the annual total by construction rather than by estimation (Frisch 1936). The link from these flows to the outstanding liability is the intertemporal accounting identity under which the sequence of balances cumulates into the stock (Obstfeld and Rogoff 1995), and the interest carried inside the outlay total is the price that equates income across dates (Fisher 1930).

The three flows are read as the gross sides of the fiscal stance, since revenue measures what the public sector withdraws from private income in the current period and expenditure measures what it returns as purchases and transfers. Their difference is the draw on national saving whose accumulation forms the outstanding public liability (Obstfeld and Rogoff 1995), so the stance is judged from both sides jointly rather than from the net figure alone (Copeland 1949). The interest embedded in the outlay total ties that stance to the intertemporal price of public income (Fisher 1930).

The receipts side is the more cyclical of the two, since tax bases move with activity while a large part of outlay is committed in advance, and this asymmetry makes national tax revenue an activity-sensitive series in the tradition of cyclical measurement (Burns and Mitchell 1946; Zarnowitz 1992). The cumulative construction sharpens the reading, because a year-to-date total compares directly with the same point of preceding years and shows whether receipts are running ahead of or behind the pace the annual budget implies.

Measuring the government as a separate account makes the sectoral withdrawal and injection direct objects of analysis (Fabricant 1940; Kuznets 1941), and the accounting frame places public receipts and outlays inside the economy-wide circular flow (Stone 1966; Stone, Champernowne, and Meade 1942). Read against activity, the two sides measure the reallocation of purchasing power among households, firms, and the state rather than the size of the state alone.

Expenditure sustained above revenue raises the accumulated liability whose ratio to activity has repeatedly risen alongside macrofinancial strain in the long-run record (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017), and elevated public liabilities condition the funding environment of the subsequent cycle (Lane and Milesi-Ferretti 2018). Gross positions carry information beyond what the net balance reveals (Obstfeld 2012), and the empirical connection between fiscal quantities and the level of short-term interest rates ties these flows to the broader macro-financial setting (Hamilton 1996).

For market participants the receipts and outlays of the central government are the inputs from which the funding need behind sovereign issuance is sized, since the gap between them over the year determines how much paper has to be sold and when. The price sensitivity of that paper to yield moves is governed by the duration relation between a bond's yield and the weighted term of its cash flows (Macaulay 1938; Fisher and Weil 1971), the curve that prices it carries the compensation for deferring income across dates (Hicks 1939), and the daily construction of that curve from traded prices is itself a measurement exercise (Gürkaynak, Sack, and Wright 2007).

The cumulative profile of revenue and expenditure imprints the seasonal pattern of the government's cash position, which passes into the money market through the timing of tax settlement and disbursement and appears as the liquidity condition an operating framework has to offset (Bartolini, Bertola, and Prati 2002; Bindseil 2004; Borio 1997). Desks that fund positions in government paper track those flows because collateral that becomes scarce around heavy settlement dates trades at a premium in secured funding (Duffie 1996), and the overnight rate is set in the market where that pressure finally lands (Hamilton 1996).

Revenue running behind and outlay running ahead of the budgeted pace signal a larger financing requirement, which markets read into supply expectations before any announcement. Persistent gaps accumulate into a liability whose level has been associated with a higher frequency of macrofinancial disorder (Schularick and Taylor 2012; Jordà, Schularick, and Taylor 2017), and that association informs the credit and term spreads at which sovereign and related exposures clear.

Because the three series are gross flows rather than a net position, they are read with the same caution the position-measurement literature raises about valuation and coverage (Gourinchas and Rey 2007a; Curcuru, Dvorak, and Warnock 2008), a caution that generalizes to every gross public magnitude (Obstfeld 2012).

KRGOVEXP is a not-seasonally-adjusted cumulative expenditure flow whose values are strictly positive, so the appropriate statistical object is its logarithm. The sample is 317 monthly observations spanning 2000-01 to 2026-05, and the unit-root triplet is fitted on the log level with a constant and trend.

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 at p = 0.146, whereas the nonparametric test of Phillips and Perron (1988) rejects it at p = 0.000 and the stationarity-null test does not reject trend stationarity at p > 0.10 (Kwiatkowski et al. 1992), so the three procedures disagree and no clean order of integration is assigned. A split of this kind between the parametric and the nonparametric leg is what the finite-sample study of Ng and Perron (2001) documents under near-unit moving-average errors, and the GLS-detrended escalation of Elliott, Rothenberg, and Stock (1996) is the reserved confirmation tool for such ambiguous readings.

Because the log level is not a clean stationary object, the portmanteau and the break search are run on the log difference, since a level portmanteau on a cumulating series reads mechanical autocorrelation and a level mean-break search spuriously segments the series (Perron 1989; Hamilton 2018; Bai and Perron 1998). The portmanteau statistic of Ljung and Box (1978), refining the form of Box and Pierce (1970), rejects the white-noise null on the log difference with Q = 342.36 at lag 12 and Q = 666.92 at lag 24, both at p = 0.000. The multiple-break procedure computed by the dynamic-programming algorithm of Bai and Perron (2003) finds no break in the mean of the log difference.

Seasonal tests are appropriate on this unadjusted monthly series, whose cumulative construction imprints a strong within-year pattern as outlays accumulate from a January reset toward a December total. The seasonal-unit-root test of Hylleberg et al. (1990), with the monthly mechanics of Beaulieu and Miron (1993), does not reject the unit roots at the seasonal frequencies with F = 3.47 and p = 0.263, while the seasonal dummies are jointly significant on the differenced series with F = 117.44 at p = 0.000 and the seasonal portmanteau rejects white noise at the seasonal lags with Q = 605.52 at p = 0.000. The seasonality is therefore reported as present but not as deterministic, since a stochastic seasonal component is not ruled out, which is the honest reading for an unadjusted cumulative fiscal series (Ghysels and Osborn 2001).