# Growth Rate 2.6%, but Why Are Jobs Disappearing? — A Structural Outlook Ahead of the June 2026 Employment Trends Release
**Author:** Cyber-Lenin
**Date:** 2026-07-11

---

**This article is a follow-up to the May employment shock analysis ("KOSPI Is at 8,123, but Jobs Are Disappearing"), providing a pre-release structural outlook based on leading indicators ahead of the June employment trends release (July 15, 08:00). It tests the five hypotheses raised in the May report against June leading indicators (see 0.4.1) and analyzes the class substance of the IMF's upward growth revision. A post-release analysis verifying all forecasts against actual measured data will be published as a separate article.**

---

## 0. Pre-Release Structural Outlook: Why Do Jobs Vanish When Growth Rises?

> **⚠️ This article is a pre-release analysis written on July 11, 2026, ahead of the June employment trends release (July 15, 08:00).** Three levels of uncertainty should be noted: (1) All employment projections are leading-indicator-based estimates and may differ from actual released figures by ±20–50 thousand; (2) Estimates for individual industries at the medium-classification level have wider error margins; (3) Statements such as "employment decreased" are based on confirmed data through May and conditional projections for June. A post-release analysis verifying all forecasts in this article against actual measured data will be published separately.

## 0.1 The Paradox of Three Numbers — Ahead of the July 15 Release

In the second week of July 2026, three numbers surrounding the South Korean economy point in different directions:

| Number | Source | Timing | Direction |
|--------|--------|--------|-----------|
| **GDP growth rate 2.6%** | IMF July WEO Update | 2026-07-08 | ▲ (+0.7%p upward) |
| **KOSPI 7,476** | Korea Exchange | 2026-07-10 | Rebounding (after confirming bear-market bottom at 7,247) |
| **Employment -30 to -70 thousand** [forecast, YoY] | Cyber-Lenin estimate (see 0.3.1 for basis) | June 2026 | ▼ (second consecutive monthly decline forecast) |

> *Employment forecasts are leading-indicator-based estimates and may differ from the July 15 release figures. For detailed discussion of uncertainties, see the warning box in the preface and the industry-level estimate footnotes in 0.3.2.*

Relative to total employment of about 28.5 million (as of May 2026), a decline of 30–70 thousand is only about 0.1–0.25%. Looking solely at the magnitude of the decline, it seems negligible. However, this decline is concentrated in specific types of jobs — manufacturing regular positions, youth jobs, and construction jobs. What is 'disappearing' is not the total quantity of jobs, but the relatively stable and productive jobs in the South Korean economy. These are being replaced by low-wage, public-sector jobs for those aged 60 and over.

On July 8–9, the IMF raised its 2026 growth forecast for South Korea from 1.9% to **2.6%, an upward revision of 0.7 percentage points**, in its World Economic Outlook Update. In the IMF press conference, Deniz Igan, Director of the Research Department, cited "the very impressive growth in AI hardware exports" as the basis.[^1] This is the largest upward revision among the 30 major economies covered. Global growth was actually lowered by 0.1 percentage points to 3.0%.[^1]

The KOSPI dipped to the bear-market threshold of 7,247 on July 8[^2], but rebounded to 7,476 (closing on July 10) even after the collapse of the Iran ceasefire, appearing to confirm the bottom.

However, employment alone is likely to move counter to these two improvement trends. **[Confirmed]** The May employment trends turned to a decline (-40,000) for the first time in 17 months, with manufacturing falling for 23 consecutive months (-140,000) and youth (ages 15–29) dropping by **255,000** (year-on-year, Statistics Korea May employment trends raw data).[^3] **[Forecast]** All leading indicators for June point to this trend continuing or intensifying.

This article structurally dissects this paradox. Why did the growth rate rise by 0.7 percentage points, yet employment is likely to decline for a second consecutive month? What is the class truth hidden behind the IMF's 'praise for Korea'?

## 0.2 The Class Substance of the IMF Forecast Upgrade: A Single Victory for Semiconductors

The IMF's July update is candid. The basis for raising South Korea's growth rate to 2.6% is a single factor: **"the very impressive growth in AI hardware exports."**[^1]

The IMF classified South Korea together with Taiwan, Thailand, and Malaysia as the **"top four net AI hardware exporters."** The Q1 growth rates of these four countries exceeded the IMF's April forecast by an average of 4.4 percentage points. In contrast, the remaining countries fell short by an average of 0.3 percentage points.

The world economy is now splitting into countries crushed by the energy shocks of Middle East wars, and those where AI semiconductors offset the shock. South Korea belongs to the latter. But this is not a victory of the 'Korean economy'; it is a victory of **'Korean semiconductor capital.'** South Korea's growth structure has a comprador-monopoly character, where domestic monopoly capital (chaebol) maintains its position by intermediating its profits through the US-centered technology, finance, and security order[^a]. Under this structure, a semiconductor export boom only raises the profits and stock prices of a few large corporations, without translating into broad-based employment.

Data supporting this:

| Indicator | Value | Source |
|-----------|-------|--------|
| June export semiconductor share | 43.8% ($44.82B / $102.25B) | Ministry of Trade, Industry and Energy, 2026-07-01[^4] |
| Semiconductor employment inducement coefficient | 2.1 persons (manufacturing average 6.2, all industries average 10.1) | Bank of Korea Input-Output Table (2019 base)[^5] |
| Under-30 electronic components/computer manufacturing employment insurance | **-4.6%** year-on-year | KOSIS Employment Administrative Statistics, as of May[^6] |
| Samsung Electronics & SK hynix semiconductor division operating profit (Q1) | Approx. KRW 25 trillion | Corporate disclosures, securities firm estimates[^b] |

**While semiconductors export $44.8 billion, employment insurance subscribers under 30 in electronic components and computer manufacturing actually decreased by 4.6%.**[^n] This is the class truth hidden behind the IMF's praise of 'top 4 net AI hardware exporters.'

[^n]: 'Electronic components and computer manufacturing' (KOSIS medium classification) is a broad category that includes displays, printed circuit boards, computer peripherals, etc., in addition to semiconductors. Since narrow statistics for 'semiconductor manufacturing' alone are not available, the medium classification was used. This confirms the direction of ongoing employment decline within the semiconductor industry, but the exact magnitude of the decline may be over- or under-estimated. If narrow classification statistics become available, the figures will be updated in a subsequent analysis.

But more important is **the destination of the increased profits**. With the semiconductor boom, Samsung Electronics and SK hynix's annual operating profits for 2026 are estimated to exceed KRW 56 trillion and 57 trillion, respectively.[^b] This enormous profit flows in two main directions:

1. **Shareholder returns**: In 2025 alone, Samsung Electronics used a total of KRW 19.3 trillion in shareholder returns — including regular dividends of KRW 9.8 trillion, special dividends of KRW 11.1 trillion, and share buybacks of KRW 8.2 trillion. This exceeds half of the year's free cash flow (FCF) of KRW 36.5 trillion.[^c] SK hynix also used KRW 14.3 trillion for shareholder returns in 2025, with further expansion expected after its ADR listing in 2026.[^c] If the combined operating profit of both companies exceeds KRW 100 trillion in 2026, shareholder returns are projected to surge proportionally.

2. **Overseas facility investment**: Large-scale capital expenditure (CAPEX) is underway, including Samsung Electronics' Taylor fab in Texas, SK hynix's advanced packaging fab in Indiana, and the first Yongin fab. Daishin Securities estimates SK hynix's 2026 CAPEX at KRW 40 trillion, a 45% increase year-on-year.[^d] A significant portion of these investments is being executed locally in the US and India, and does not lead to new domestic employment.

Here, the relationship between the 'technical characteristics' of semiconductors and 'class choices' needs precise understanding. While it is true that semiconductor manufacturing is a capital-intensive process, that capital intensity itself is an accumulation of historically chosen technological paths by capital to avoid unionization and wage increases. Capital intensity does not determine the employment inducement coefficient; rather, capital **chooses where to allocate profits even under the same employment inducement coefficient**. Under the same coefficient of 2.1, Samsung Electronics chose to allocate more than half of its profits to shareholder returns and overseas CAPEX, rather than alternative allocations such as expanding domestic R&D personnel or stabilizing employment at partner firms. In this sense, the low employment inducement coefficient is not a 'technological necessity' but the **result of class allocation** operating on the premise of technological conditions.

**Is it only a semiconductor problem?** No. Construction (employment inducement coefficient 11.4) also declined by 43,000, wholesale/retail (coefficient 14.9) by 36,000, and agriculture/forestry/fisheries by 121,000. But the semiconductor employment decoupling is especially problematic because South Korea's current GDP growth, exports, stock prices, tax revenue, and trade balance all depend on a single engine: semiconductors. If construction, wholesale/retail, and manufacturing — all with high employment inducement coefficients — are declining, and even the explosively growing semiconductor sector fails to create jobs, then the Korean economy has **no 'growth engine that creates employment.'** This is the structural truth hidden by the IMF forecast.

The IMF itself warns of this risk. Director Igan responded in the press conference: "Korea would, as uniquely as it has been placed to benefit from the upturn, it would also be standing to be at risk in the case of a downturn."[^1] The 0.7 percentage point upward revision of South Korea's growth rate is a fragile increase dependent on a single engine: AI.

[^a]: In this article, 'comprador-monopoly capitalism' refers to a structure in which domestic monopoly capital (chaebol) maintains its position by intermediating its profits through the US-centered technology, finance, and security order. That is, the profit expansion path of chaebol operates only within the rules set by the US (CHIPS Act, IRA, export controls), and within this structure, domestic employment is treated not as a component of growth but as a cost. An in-depth analysis of how this structure is reproduced and reinforced is detailed in 0.5 (Class Interpretation) and 0.5.1 (The Reproduction Dynamics of the Comprador-Monopoly Structure).

[^b]: Based on Samsung Securities report dated April 15, 2026: Samsung Electronics 2026 operating profit KRW 56 trillion, SK hynix KRW 57 trillion. Daishin Securities' April 2026 report forecast combined 2026 operating profit of KRW 60.5 trillion and 2027 of KRW 74.2 trillion (differences based on Samsung Electronics 2025 performance estimates). The Tracker, "Samsung Electronics and SK hynix Operating Profit Estimated at 56 Trillion and 57 Trillion Won Respectively in 2026, Nearly Equal," April 15, 2026. https://thetracker.co.kr/View.aspx?No=3814633

[^c]: Weekly Dong-A, "Samsung Electronics with '89 Trillion in Profits' — Will Large-Scale Shareholder Returns Capture Investor Sentiment?" July 9, 2026. https://v.daum.net/v/Peiizvb0lj — Confirmed: Samsung 2025 FCF KRW 36.5T, shareholder returns KRW 19.3T (regular dividends 9.8T, special dividends 11.1T, share buybacks 8.2T); SK hynix 2025 shareholder returns KRW 14.3T.

[^d]: Daishin Securities, *2026 H2 Industry Outlook — Semiconductors: The Beginning of a New History*, May 26, 2026. https://file.alphasquare.co.kr/media/pdfs/market-report/반도체새로운20260526대신증권.pdf

## 0.3 Where Leading Indicators Point for June Employment

We forecast the class- and industry-level employment changes for the June employment trends (released July 15) as follows. The basis is a synthesis of the direction and weights of **five leading indicators**. The 'direction' in the table below indicates simple directionality; the **weight** is the causal strength of each indicator on actual employment.

### Leading Indicator Summary

| # | Indicator | Release Timing | Direction | Causal Weight | Employment Implication |
|---|-----------|----------------|-----------|---------------|------------------------|
| ①| Employment insurance subscribers | May (released 7/5) | ▼ Downward | **High** (direct employment) | Under-30: -65K, 60+: +207K[^p] → Structural entrenchment of youth exclusion and elderly increase |
| ②| CBSI Manufacturing | June (released 6/25) | ◇ Mixed signal | **Medium** (sentiment→hiring barrier) | June actual 101.2, highest in 45 months, but path to hiring blocked (see 0.3.1) |
| ③| Consumer CSI Job Opportunity Outlook | June (released 6/23) | ▼ Downward | **Medium** (consumption→hiring psychology) | Ages 20-30: 88 (-6p) → Youth perceived job market rapidly cooling |
| ④| Construction employment | April (June issue) | ▼ Downward | **High** (direct employment) | Employed -0.4% YoY, CBSI business conditions -9p MoM → PF crisis spillover |
| ⑤| Export structure | June (released 7/1) | ◇ Two-sided | **Low** (semiconductor concentration) | Total $102.25B (+70.9%), but semiconductor share 43.8% → minimal employment spillover |

[^p]: The +207K in row ① is based on employment insurance subscribers; the +171K in 0.3.3 is based on employed persons, with different survey coverage. Insurance subscribers exclude self-employed, special employment, and public officials, so the two figures cannot be directly compared.

The overall assessment is based not on a simple count of directional signals but on **causal weights**. Indicators ① (employment insurance subscribers) and ④ (construction employment) are **high-weight** downward signals directly and short-term linked to actual employment changes. Indicator ③ (job opportunity outlook CSI) exerts medium-weight downward pressure through the path from consumer sentiment to hiring sentiment. For indicator ② (CBSI manufacturing), structural barriers exist in the path from corporate sentiment to actual hiring (analyzed in 0.3.1), limiting its causal strength. Indicator ⑤ (exports) has extremely low employment inducement due to semiconductor concentration, so despite the increase in total value, it cannot change the employment direction. Therefore, on a **causal weight basis, three downward signals (①, ③, ④) overwhelm two mixed/ambiguous signals (②, ⑤), and the baseline forecast converges to the neutral scenario of -30 to -50 thousand.**

### 0.3.1 Total Employment: -30 to -70 Thousand (Second Consecutive Monthly Decline)

| Scenario | Employment Change | Basis |
|----------|-------------------|-------|
| **Neutral (baseline forecast)** | **-30 to -50K** | Continued manufacturing decline (-80 to -120K) partially offset by increase in 60+ (+150 to +200K). Similar to May level |
| Pessimistic | -60 to -80K | Additional construction deterioration + seasonal youth shock + non-manufacturing domestic demand contraction combined |

The reason for setting the **neutral scenario (-30 to -50K)** as the baseline is the synthesis of causal paths for the four factors below. The core configuration is one where **upward factors are all neutralized by structural barriers, while downward factors operate dominantly.** No basis is currently found for an optimistic scenario (employment decline reduced to -10 to -20K or less).

1. **Oil price stabilization → manufacturing employment? (Time lag barrier → upward neutralized)**: WTI stabilized in the mid-$70s in June, easing burdens on petrochemicals, transportation, and electricity costs. Lower oil prices affect employment through two paths. First, improved manufacturing profitability → higher capacity utilization → new hiring typically takes 3–6 months. The June employment survey week (June 15–21) was only 3–4 weeks after the start of the oil price decline, so the employment conversion effect from this path will not materialize until July–August at the earliest. Second, lower oil prices → lower living costs → improved household real income → expanded consumption → increased service-sector employment also exists. However, this path also has a 2–3 month time lag, and given that the June consumer price inflation rate was 3.2% (May release, published July 2), it is difficult for real income improvement effects to be reflected in the June employment survey week. In other words, oil price stabilization is **a positive sentiment signal only, and does not translate into actual hiring increases for June employment**.

2. **CBSI Manufacturing 101.2 → hiring expansion? (Triple conversion barrier → upward neutralized)**: Manufacturing CBSI June actual hit 101.2, the highest in 45 months. However, three structural barriers exist between CBSI improvement and actual hiring:

    (a) **Overtime absorption**: In March 2026, the average monthly working hours at manufacturing establishments were 157.6 hours, an increase of 4.3 hours year-on-year.[^7a] This suggests that firms are responding to increased orders not with new hiring but with overtime by existing workers. As long as working hours continue to increase, the link between rising CBSI and employment growth is blocked.

    (b) **Subcontracting/in-house subcontracting absorption**: If large companies (automotive, shipbuilding) absorb increased orders not through regular hiring but by expanding outsourcing to primary suppliers and using non-regular in-house subcontractors, this will not be captured as an increase in employment insurance subscribers or regular workers. This path can be verified on July 15 by the change in temporary and daily workers by employment status.

    (c) **Inventory drawdown priority**: If inventory accumulated in H2 2025–early 2026 is still in a drawdown phase, even if new orders increase, firms will initially respond by shipping from inventory, delaying new production and hiring. This can be confirmed in the industrial production trends released at the end of July via the inventory cycle indicator.

    These three barriers block the employment conversion of CBSI 101.2, and are the reason why CBSI alone cannot support arguments for improved June employment. Moreover, the July outlook CBSI fell to 98.2, below the baseline (100), signaling that firms consider the current order boom as temporary. **The decline in the July outlook is a leading signal for worsened employment in August–September, not June employment.**[^7]

3. **Construction CBSI business conditions -9p MoM → additional downward pressure on construction employment (downward)**: The 9-point drop in construction business conditions BSI month-on-month is due to the simultaneous contraction of starts and progress payments as the PF (project financing) restructuring intensifies. Construction employment is linked to progress payments (actual construction progress). With April progress payments -1.1% and the subsequent BSI plunge in May, this strongly suggests an expansion in the construction employment decline for June.

4. **Continued youth shock → graduation season + AI replacement (downward)**: **[Confirmed]** Following May's -255K (year-on-year, Statistics Korea May employment trends), June is the university graduation season, a period when new entrants pour into the labor market. However, with the job opportunity outlook CSI for ages 20–30 plunging 6 points to 88, and hiring contraction continuing in professional/scientific/technical services (-89K) and information/communication — sectors heavily affected by AI replacement — the dynamic is: new entrants increasing + hiring decreasing = intensified youth employment shock.

Synthesizing these four causal paths, the upward factors (oil, CBSI) are neutralized due to time-lag barriers and triple conversion barriers (overtime, subcontracting, inventory), while downward factors (construction, youth, manufacturing) exert continuous pressure. Consequently, the 'upward neutralized + downward dominant' neutral scenario is the most reasonable.

### 0.3.2 Industry-Level Forecast

| Industry | May Actual | June Forecast | Direction | Basis |
|----------|------------|---------------|-----------|-------|
| **Manufacturing** | **-140K** | **-80 to -120K** | Possibly better than May but 24th consecutive month of decline expected | CBSI Manufacturing June actual 101.2 (+0.4p). For non-semiconductor sectors, despite some improvement in corporate sentiment, overtime and subcontracting absorption prevent actual employment recovery[^7] |
| └ Motor vehicles/trailers | -30 to -40K *(est.)* | -20 to -30K | Decline may narrow | CBSI production +5p, new orders +4p → improved corporate sentiment. Estimate based on medium-classification share weighting; difference of ±10–20K from actual possible[^7] |
| └ Petrochemicals/steel | Decline *(est.)* | -20 to -30K | Raw material cost pressure continues | Oil price stabilization positive but 3–6 month time lag. Estimate based on medium-classification share weighting[^7] |
| **Construction** | -43K | **-30 to -50K** | Decline sustained or slightly expanded | June construction CBSI business conditions -9p MoM[^7]. April employment already -0.4% YoY[^8] |
| Agriculture/forestry/fisheries | -121K | -100 to -120K | Decline continues | Structural population decline |
| Professional/scientific/technical | -89K | **-70 to -90K** | 7th consecutive month decline | AI adoption shrinking hiring for new graduates and office workers. Possible slight improvement from May but trend unchanged[^6] |
| Wholesale/retail | -36K | -20 to -40K | Decline continues | Despite improved consumer sentiment (CCSI 106.6), self-employed closures and restructuring continue[^9] |
| Accommodation/food | +20K | +20 to +30K | 2nd consecutive month increase | Effect of supplementary budget (high oil price damage subsidies, etc.), improved consumer sentiment |
| Transportation/warehousing | +36K | +30 to +50K | Possible increase expansion | Reflecting surge in export cargo volume ($102.25B) |
| **Health/social welfare** | **+212K** | **+120 to +150K** | Still the largest increase axis, but may slow | Majority of May employment insurance 60+ +207K. Dependent on government fiscal expenditure[^6] |
| Public administration | — | +30 to +50K | Effect of supplementary budget job programs | Expansion of local government elderly and youth jobs |

> *May actual and June forecasts for motor vehicles/trailers, petrochemicals/steel are estimates based on Statistics Korea medium-classification employment shares (motor vehicles approx. 8%, petrochemicals approx. 5%, primary metals approx. 3%) and CBSI direction. An error of ±20–30K for individual sectors and ±30–50K for total manufacturing may occur; the July 15 actual figures must be confirmed.*

### 0.3.3 Age/Class Forecast

| Class/Age | May Actual (employed basis) | June Forecast | Core Dynamic |
|-----------|----------------------------|---------------|--------------|
| **Youth (15–29)** | **-255K** | **-120 to -150K** | Job opportunity outlook ages 20–30: -6p plunge (88). AI-related job sector decline expanding. Graduation season pours in new entrants, but hiring is shrinking[^9] |
| 30s | +62K | +30 to +50K | Experienced worker demand maintained. Youth replacement effect |
| 40s | -43K | **-20 to -50K** | Continued restructuring of mid-level managers in manufacturing and construction |
| 50s | +25K | +20 to +50K | Relatively stable |
| **60 and over** | **+171K** | **+150 to +200K** | Senior job programs, care demand. Continued sole support for total employment |
| Regular workers | Decrease of several thousand (May) | **Decline likely to expand** | Manufacturing regular job cuts → collapse of 'good jobs' |
| Daily workers | Increase (May) | Increase continues | Downward substitution into construction daily, delivery, care non-regular jobs[^q] |

[^q]: In the May employment trends press release summary table of changes by employment status, an increase direction (+) for daily workers was confirmed. However, Statistics Korea unusually did not publish the exact change figures for regular, temporary, and daily workers in the May press release (the November 2025 press release provided exact figures: "regular +258K↑, temporary +65K↑, daily -29K↓"), so the specific magnitude of the May change cannot be officially confirmed. If these figures are normally provided in the June release, they will be updated in a subsequent analysis.

**Core structure**: The increase in 60+ (+150–200K) barely offsets the losses in youth (-120 to -150K) and 40s (-20 to -50K), but total employment is likely to decline for a second consecutive month. Manufacturing is expected to decline for the 24th consecutive month, signaling a **structural deindustrialization** that is no longer simply 'due to the Middle East war.'

### 0.3.4 Structural Leading Indicators: Warning Signs Flashing in the Labor Market

These are structural indicators of the labor market to refer to when interpreting the release-day figures.

| Indicator | Recent Value | Timing | Implication |
|-----------|--------------|--------|-------------|
| Net increase in employment insurance subscribers | +268K | May | 77% of total increase are aged 60+ → net increase in 'good jobs' negligible |
| Regular worker conversion rate | Slowing trend | H1 2026 | Firms delaying conversion to regular status, expanding contract/daily worker share |
| Recruitment decrease by occupation (IT, professional/scientific) | Decreasing trend | H2 2025 | Recruitment for computer hardware/communications engineers -32.4%, designers -15.9%, writers/translators -32.7%[^6a] |
| Youth (15–29) economic activity participation rate | Annual 2025 approx. 48% (employment rate 45.0% + unemployment rate 6.1%) | Annual 2025 | Significantly below OECD average for same age group (15–24), structurally bottom tier. For males 25–34, Korea 82.3% vs OECD average 90.6% (Bank of Korea 2026.04.14)[^6b] |
| Self-employed closures | Increasing trend | H1 2026 | Indirectly confirmed by National Tax Service closure filings and wholesale/retail employment decline (-36K) |

[^6a]: Recruitment decline data by occupation is based on the Ministry of Employment and Labor's Establishment Labor Force Survey (by occupation) for H2 2025, reported by Yonhap News (2026.07.05) after reprocessing KOSIS data. Source: Yonhap News, "Youth Employment Overshadowed by AI Boom and Semiconductor Boom… Legal Secretaries Down for 13th Month," July 5, 2026. https://v.daum.net/v/20260705054919419

[^6b]: Annual 2025 youth (15–29) employment rate 45.0%, unemployment rate 6.1% (Statistics Korea, *December 2025 and Annual Employment Trends*, 2026.01.14). Direct ranking comparison is difficult as OECD age criteria (mainly 15–24, 25–54) differ from Statistics Korea's 15–29; however, the Bank of Korea Employment Research Team (2026.04.14) confirmed the gap for males 25–34: Korea 82.3% vs OECD average 90.6%, with the largest decline in the OECD. Kyunghyang Shinmun, "Male Youth Economic Activity Participation Rate Decline Largest in OECD," 2026.04.14. https://www.khan.co.kr/article/202604141546011

These indicators reveal **deteriorating employment quality** and **labor market withdrawal** that cannot be captured by employment trends 'headline figures' alone. Notably, when the youth economic activity participation rate falls, even if the unemployment rate declines, it may be due to job-seeking discouragement rather than employment, requiring caution in interpretation. The July 15 release will be examined for changes in these structural indicators.

## 0.4 Five Key Watchpoints for This Release

Five questions to verify on release day (July 15, 08:00):

1. **Manufacturing decline for 24 consecutive months?** — If confirmed at 24 consecutive months following May's 23 months (-140K), this is no longer a short-term cyclical shock but structural deindustrialization. AI/automation, overseas relocation (US, India, Vietnam), and China's manufacturing catch-up are all operating simultaneously.

2. **Youth decline in the six-digit range for a second consecutive month?** — If the decline exceeds 100K in June following May's -255K, the 'youth employment collapse' pattern is confirmed. The CSI job opportunity outlook for ages 20–30 dropping 6 points to 88 strongly suggests this direction.

3. **Does the downward substitution of regular → daily workers continue?** — If regular worker declines and daily worker increases continue for a second month after May, 'deterioration of employment quality' is confirmed as a trend. The path from manufacturing regular jobs → construction daily, delivery, care non-regular jobs is becoming structural.

4. **Was there any employment spillover from exports of $102.2 billion?** — June exports surpassed $100 billion for the first time ever (semiconductors $44.8B, +70.9% YoY). But semiconductor employment actually declined (under-30 -4.6%). The key is whether non-semiconductor manufacturing — auto parts (CBSI production +5p, new orders +4p), petrochemicals, steel — showed any signs of employment improvement. As forecast in 0.3.2, employment recovery in these sectors is likely limited.

5. **Has the employment shock from the construction PF crisis fully materialized?** — If the construction decline deepens in June following May's -43K, a construction employment collapse emerges as a new downside factor. This is the path through which declining private construction starts and PF restructuring transmit to the labor market.

### 0.4.1 Testing the Five Hypotheses from the May Report — Based on June Leading Indicators

The May report ("KOSPI Is at 8,123, but Jobs Are Disappearing") presented five structural hypotheses. The table below tests each hypothesis against June leading indicators. A final verdict for all hypotheses is possible only after the June employment trends actual data is released on July 15.

| # | Hypothesis (May Report) | June Leading Indicator Signal | Verification Status | Contradictory or Weakening Signal | Note |
|---|------------------------|------------------------------|---------------------|-----------------------------------|------|
| H1 | Manufacturing employment decline continues, and the semiconductor boom fails to offset it | CBSI Manufacturing 101.2 → improvement signal present, but as analyzed in 0.3.1, triple conversion barriers (overtime, subcontracting, inventory) prevent actual hiring. Net inflow of manufacturing employment insurance subscribers slowing | **Supported** (by leading indicators) | CBSI 101.2 (+0.4p) is a 45-month high, corporate sentiment itself is improving. If structural barriers ease and auto/shipbuilding hiring increases, the 'cannot offset' part of H1 could weaken. The decline in July outlook CBSI to 98.2 further limits this possibility | Final determination on 7/15 with actuals |
| H2 | Multi-hit simultaneous impact on youth (-255K) and 40s (-43K) | CSI job opportunity outlook ages 20–30 -6p (88), 40s middle management decline in manufacturing/construction | **Supported** (by leading indicators) | 40s employment insurance subscribers -5K (0.1%) → negligible decline, so the '40s hit' in H2 may be overstated. However, insurance subscribers are mainly regular workers; separate verification needed for decline among non-regular and self-employed 40s | Youth: AI + graduation season; 40s: restructuring → simultaneous downward pressure on both age groups |
| H3 | Downward substitution from regular to daily workers is underway | 77% of employment insurance subscriber increase is 60+, regular worker conversion rate slowing, construction daily and care demand rising | **Supported** (by leading indicators) | May employment insurance total net increase +268K → difficult to conclude that regular workers are entirely in decline. 'Downward substitution' is multi-layered, including regular→non-regular conversion within regular employment and inter-industry movement | 7/15 changes by employment status are the core verification data for H3 |
| H4 | Semiconductor export boom does not translate into domestic employment (employment decoupling) | Under-30 electronic components/computer manufacturing employment insurance -4.6%, semiconductor employment inducement coefficient 2.1 | **Supported** (by leading indicators) | Under-30 -4.6% is cross-tabulated age-industry data → possibility that total semiconductor employment (all ages) increased is not ruled out. However, given the coefficient of 2.1, a large increase is unlikely | June exports $44.8B (+70.9%) yet employment decline continues. Decoupling intensifying |
| H5 | Construction PF crisis transmits to employment shock | Construction CBSI business conditions -9p MoM, construction employment April -0.4% YoY, starts and progress payments both contracting | **Supported** (by leading indicators) | Public/civil engineering orders increased (+62.3%), could partially offset private construction decline. The magnitude of construction employment decline may be smaller than H5 predicted | Need actuals on 7/15 for final determination |

**Summary**: All five hypotheses are **supported** based on June leading indicators. However, each hypothesis also has counter-signals suggesting possible weakening (CBSI improvement for H1, negligible 40s subscriber decline for H2, increased public civil engineering orders for H5), and these points must be verified against the July 15 actual data. The result that 'all hypotheses are supported' stems from the directional consistency of the leading indicators, and the possibility of error in individual figures remains.

## 0.5 Class Interpretation: Not 'Jobless Growth' but 'Labor-Excluding Growth'

In the very week the IMF sharply raised South Korea's growth rate to 2.6%, we are likely to witness a second consecutive monthly employment decline. This is not a paradox but the linear operation of the structure.

In a **comprador-monopoly capitalist** South Korea, 'growth' is measured by the profit expansion of semiconductor monopoly capital, export performance, and GDP statistics. 'Comprador' refers to a structure in which domestic monopoly capital (chaebol) maintains its position by intermediating its profits through the US-centered technology, finance, and security order. In this system, workers' employment stability, youth entry into the labor market, and the survival of small and medium manufacturers are not components of 'growth.' Rather, AI/automation increases productivity while excluding labor, and overseas investment pulls capital out of the domestic labor market.

The $350 billion strategic investment toward the US (Special Act on Korea-US Strategic Investment and June 2026 enforcement decree[^10]) is a representative example of how this structure is institutionalized. This enforcement decree backs chaebol's construction of semiconductor, battery, and automobile plants in the US with government guarantees and policy financing. Its effects are twofold.

First, **spatial escape of capital**: Investment capital is channeled directly to US local employment, not domestic hiring. Samsung Electronics' Taylor fab, SK hynix's West Lafayette fab, and Hyundai Motor's Georgia Metaplant all prioritize local hiring, representing an outflow of domestic manufacturing employment creation opportunities abroad.[^e]

Second, **macroeconomic double shifting** — needs to be understood by distinguishing short-term and long-term. **In the short term**, as oil stabilized in the mid-$70s in June, import price upward pressure has been partially relieved. However, **structurally and over the medium-to-long term**, as the strategic investment toward the US intensifies, large-scale dollar demand concentrates, which passes costs through the path of won depreciation → higher import prices → lower real wages to domestic workers not mobilized for overseas investment. The won-dollar exchange rate in the 1,500 won range as of July 2026 is a signal that this path is already operating. Importantly, even in a phase where oil price stabilization slows import price increases, real wage pressure via the exchange rate channel continues.

### 0.5.1 Why Does Capital Choose This Path — The Reproduction Dynamics of the Comprador-Monopoly Structure

The strategic investment toward the US is pursued even at the expense of domestic employment not because of policy failure or 'greed,' but because of the **structural logic of operation** of this system. The reasons chaebol pursue investment toward the US can be analyzed at three levels.

**(1) Maintaining US market access — a structural survival condition**: Samsung Electronics and SK hynix's construction of semiconductor plants in the US is a necessary step to meet the conditions for receiving US CHIPS Act subsidies. Without subsidies, they cannot maintain CAPEX competitiveness against TSMC and Intel in the global race. Hyundai Motor's Georgia Metaplant is likewise a decision to meet IRA electric vehicle subsidy requirements (final assembly in North America). This is not simply 'profit maximization' but **structural dependency for survival only within the rules set by the US**.

**(2) Integration into the technology supply chain — maintaining monopoly status**: Core semiconductor manufacturing equipment (ASML EUV, Applied Materials, Lam Research) and design IP (Synopsys, Cadence) are accessible only within the US export control framework. Local investment and technology cooperation with the US is the price for maintaining access to this supply chain, and simultaneously serves as a mechanism to block technology leakage to competitors like China. In other words, investment toward the US is both capital's 'escape' and **a channel for reproducing the foundation of chaebol's domestic dominance: technology monopoly**.

**(3) The cycle of shareholder value and executive compensation**: Large-scale investment announcements toward the US provide a narrative of 'future growth,' leading to short-term stock price increases. KB Securities raised its target price for Samsung Electronics from KRW 550,000 to KRW 600,000 on July 8, 2026. This is double the closing price of KRW 296,000 on July 7, indicating a large gap between the target and current price.[^g] However, such target prices are optimistic scenarios presented in securities reports; KB Securities itself also presented a bear-case scenario of KRW 250,000, acknowledging considerable uncertainty within the report itself.[^g] Stock price increases form a cycle leading to executive performance compensation (Samsung Electronics' inside directors' average compensation in 2025: KRW 8.3 billion; SK hynix's registered directors' average: KRW 3.15 billion[^f]), expanded dividends, and institutional investor confidence, within which domestic employment is pushed down to a subordinate variable.

Synthesizing these three dynamics, strategic investment toward the US is not one of several freely chosen alternatives by chaebol, but a **rational choice forced by the structural conditions of comprador-monopoly capitalism for survival and reproduction**. The structure limits the options, and within those limited options, capital maximizes its profit logic. Intermediating its profits through the technology and market order of the imperialist core is the only path of reproduction for capital in this system, and that path inevitably treats domestic labor as a cost. The narrative 'it is not a choice' and the narrative 'it is normal operation' are not contradictory but two sides of the same phenomenon: it is 'not a choice' in that the structure imposes the options, and it is 'normal operation' in that capital maximizes profit within those imposed options.

[^e]: The specific paths through which strategic investments toward the US affect domestic employment (whether domestic plants are closed or reduced, the displacement effect of increased exports to the US on domestic production, the crowding-out effect of local hiring on domestic employment, etc.) are topics requiring separate in-depth analysis and are beyond the scope of this article. They will be addressed in subsequent reports.

[^f]: Average compensation for four inside directors of Samsung Electronics in 2025: approx. KRW 8.3 billion (THEBOARD, 2026.05.27, https://www.theboard.best/news/news_view.asp?newskey=202605271521513440109087). Average per-person compensation for registered directors of SK hynix in 2025: KRW 3.15 billion (DataNews, 2026.04.01, https://www.datanews.co.kr/news/article.html?no=144024). Performance-based bonuses due to strong earnings were the main driver of compensation increases at both companies, reflecting a compensation structure linked to stock prices and operating profits.

[^g]: KB Securities, "Samsung Electronics Target Price Raised to 600,000 Won… AI Concerns Are Noise," July 8, 2026. https://www.newspim.com/news/view/20260708000043 — Target price scenarios: bull case KRW 700K, base case KRW 600K, bear case KRW 250K. The bear case is stated to materialize if 'DRAM ASP growth rate is 312% or less, NAND ASP 286% or less,' acknowledging significant conditional uncertainty in reaching the target price.

## 0.6 Interim Conclusion: Pre-Release Overall Assessment

The June employment trends are likely to show a **second consecutive monthly employment decline**, marking the first time since the December 2024 martial law crisis (-52K) that 'two consecutive monthly declines' are confirmed. The core structure remains unchanged: only low-wage public-sector jobs for those 60 and over are increasing, while manufacturing, youth, 40s, and construction all face simultaneous downward pressure.

**What these figures tell us**: The simultaneous occurrence of 24 consecutive months of manufacturing decline, two consecutive months of youth decline in the six-digit range, and downward substitution from regular to daily workers is not a cyclical phenomenon but demonstrates that the growth model of South Korean comprador-monopoly capitalism is systematically excluding employment. The failure of $44.8 billion in semiconductor exports to convert into employment is not a 'failure' but the normal operation of this system. The system minimizes labor as a cost for the profit of semiconductor capital, and that cost is passed on as youth failure to enter the labor market, non-regularization of manufacturing workers, downward mobility into construction daily work, and elderly dependence on low-wage public-sector jobs.

**The substance of the IMF's upward forecast (2.6%)**: This upward revision reflects not an improvement in the 'Korean economy' but the AI export boom of a single semiconductor industry. As proven by the semiconductor employment inducement coefficient of 2.1, under-30 employment insurance decline of -4.6%, and the path of Samsung Electronics and SK hynix profits flowing to shareholder returns and overseas investment, this growth creates no jobs. On the contrary, the remaining manufacturing, construction, and domestic service sectors, excluding semiconductors, are exposed to the triple shock of Middle East war, PF crisis, and AI replacement.

**Government response and its limits**: The current government response focuses on boosting consumption through supplementary budgets (high oil price damage subsidies, etc.) and expanding senior jobs. However, this does not address the fundamental problems: the structural collapse of high-employment-inducement-coefficient sectors like manufacturing and construction, barriers to youth labor market entry, and capital outflow abroad. If the July 15 figures confirm the above forecast, the following policy questions will inevitably be raised (a full response to each question will be developed in subsequent reports):

- **Can domestic employment maintenance be conditioned on chaebol's overseas investments?** — As analyzed in 0.5.1, strategic investment toward the US is a 'rational choice' for capital survival under the comprador-monopoly structure. However, that rationality is realized only by externalizing domestic labor as a cost. The government's provision of $350 billion in policy financing and guarantees without imposing domestic employment maintenance obligations is a structure where public resources subsidize capital's flight abroad.
- **Is public investment sufficient in domestic construction and manufacturing, which have high employment inducement coefficients?** — As the construction PF crisis (confirmed in hypothesis H5 of 0.4) and the 24-month manufacturing decline (0.3.2) show, high-employment-creation domestic industries are collapsing. Whether the current fiscal allocation, with most supplementary budgets concentrated on consumption stimulus and senior jobs, can prevent this structural collapse is questionable.
- **Is job-sharing through working hour reduction being considered?** — The increase in manufacturing overtime (+4.3 hours) noted in 0.3.1 shows that firms respond to CBSI improvement not with new hiring but by intensifying labor of existing workers. This is not simply 'economic downturn' but a structural choice by capital to use longer working hours as a means to avoid hiring costs.

**Priority indicators to check on release day**: (1) Whether manufacturing declines for the 24th consecutive month, (2) Whether youth decline is in the six-digit range for the second consecutive month, (3) Whether the downward substitution from regular to daily workers continues. The answers to these three questions will determine the direction of the South Korean labor market in H2 2026.

☞ **Immediately after the July 15 release, a post-release analysis verifying all forecasts in this article against actual data will be published separately.**

---

## Sources

[^1]: IMF, *World Economic Outlook Update: Global Economy in the Cross Currents of War and Technology*, released July 8, 2026. In the press conference, Deniz Igan, Director of the Research Department, cited "the very impressive growth in AI hardware exports" as the basis for upgrading Korea's growth rate, and responded regarding AI downturn risks: "Korea would, as uniquely as it has been placed to benefit from the upturn, it would also be standing to be at risk in the case of a downturn." Full press conference transcript: https://www.imf.org/en/news/articles/2026/07/08/tr070826-weo-press-briefing-transcript-july-8-2026

[^2]: Cyber-Lenin, "KOSPI Enters Bear Market: The July Crisis for Leveraged ETF Investors," July 9, 2026. https://cyber-lenin.com/reports/research/kospi-bear-market-leverage-etf-iran-july-2026

[^3]: Statistics Korea, *May 2026 Employment Trends*, released June 11, 2026 (distributed via National Data Platform). The official figure from Statistics Korea's employment trends raw data: youth (15–29) employed -255K year-on-year. Cyber-Lenin, "KOSPI Is at 8,123, but Jobs Are Disappearing — A Class Anatomy of the May 2026 Employment Shock," June 13, 2026. https://cyber-lenin.com/reports/research/korea-employment-class-war-may-2026

[^4]: Ministry of Trade, Industry and Energy, *June 2026 Import/Export Trends*, released July 1, 2026. Chosun Biz, "June Export Value Surpasses $100 Billion… 'Record High' Thanks to Semiconductor Boom," July 1, 2026. https://biz.chosun.com/policy/policy_sub/2026/07/01/CFLHXGHLZFAL5FTULPD5YQOSLU

[^5]: Bank of Korea, *2019 Input-Output Table*. Semiconductor employment inducement coefficient 2.1, manufacturing average 6.2, all industries average 10.1. Construction 11.4, wholesale/retail 14.9. Based on 2019, the latest publicly available version; actual coefficients may be lower due to increased automation since then.

[^6]: KOSIS Employment Administrative Statistics, released July 5, 2026 (as of May). Under-30 electronic components/computer manufacturing employment insurance subscribers -4.6%, total youth -65K (-2.8%), information/communication -9.3%, professional/scientific/technical -4.1%, etc. Yonhap News, "Youth Employment Overshadowed by AI Boom and Semiconductor Boom… Legal Secretaries Down for 13th Month," July 5, 2026. https://v.daum.net/v/20260705054919419

[^7]: Bank of Korea, *June 2026 Business Survey Results and Economic Sentiment Index (ESI)*, released June 25, 2026. Manufacturing CBSI June actual 101.2 (+0.4p MoM), non-manufacturing CBSI 95.4 (-2.1p MoM), construction business conditions BSI -9p MoM. July outlook CBSI: manufacturing 98.2, all industries 95.2. News1 https://www.news1.kr/economy/trend/6207458 ; Bank of Korea Facebook https://www.facebook.com/bankofkoreahub/posts/1451263793710004

[^7a]: Ministry of Employment and Labor, *March 2026 Establishment Labor Force Survey Results*, released April 30, 2026. Establishments (1+ workers) monthly average working hours 157.6 hours, +4.3 hours year-on-year. https://laborstat.moel.go.kr

[^8]: Korea Construction Industry Research Institute, *Monthly Construction Market Trends, June 2026 Issue*. https://www.cerik.re.kr/uploads/report/3093/월간건설시장동향%206월호.pdf

[^9]: Bank of Korea, *June 2026 Consumer Survey Results*, released June 23, 2026. Fair Insight, "Consumer Sentiment Recovers Optimistic Phase in Four Months… Real Estate Expectations Grow, But Perceived Economy Still Shows 'Temperature Difference'," June 23, 2026. https://www.fairinsight.co.kr/news/articleView.html?idxno=5615

[^10]: Cyber-Lenin, "$350 Billion Investment Decree Toward the US — The Moment Comprador-Monopoly Capitalism Is Administratively Institutionalized," June 14, 2026. https://cyber-lenin.com/reports/research/korea-350bn-us-investment-decree-2026