Growth for Whom?

For Cici Lv, a Shenzhen salesperson selling electric-bicycle batteries, winning an order could mean very little commission. Reuters’ January 20 report describes months of messages with one client ending in a purchase of one battery and a commission below US$2. It is one worker’s account of one order, not her monthly pay or an estimate for exporters generally.

Reuters interviewed 14 salespeople about the preceding export cycle. The inspected text is its attributed MarketScreener syndication; Reuters’ own page was inaccessible. This earlier case identifies a question about how orders become earnings. It does not measure household experience during the January–September 2026 period covered by the releases below.

China’s official preliminary accounts report 4.6 percent real GDP growth across the first three quarters of 2026. The overview reports 1.1 percent growth in consumer-goods retail sales. One describes aggregate production at constant prices; the other describes a retail category. Neither directly measures household confidence.

Lv’s account does not invalidate an aggregate result. It shows why an argument about household gains needs another step: evidence connecting production and sales to earnings, hours and security. That connection cannot be supplied by placing an earlier interview beside a later growth rate.

Goods-and-services retail growth is 2.6 percent, a broader category than consumer-goods retail. Investment excluding rural households falls 7.2 percent; excluding real estate changes the scope and gives a 4.2 percent decline. Per-capita disposable income has separate nominal and real growth figures, 5.3 and 4.4 percent. Each comparison has to retain its definition.

The dedicated household release adds earnings-related evidence for that same January–September period. NBS reports wage income per resident of RMB 19,653, up 5.3 percent in nominal terms. Median per-capita disposable income is RMB 28,388, up 4.6 percent nominally. Its weighted survey covers 160,000 households. Neither figure is an employee’s average salary; the median is defined by surveyed households’ per-capita income.

This narrows the gap between aggregate production and household resources. It does not identify which employers or sectors supplied the gains, track the same workers’ pay, or establish changes in hours. The figure keeps population and price basis visible rather than combining unlike measures into a score.

The October 27 industrial-profit release adds a distribution question inside production. For January–September, NBS reports aggregate profit growth of 15.5 percent, electronics manufacturing profit growth of 103.8 percent and an automobile manufacturing decline of 22.8 percent. It covers industrial legal entities with annual main-business revenue of at least RMB 20 million. These are current-price profit measures on a comparable reporting scope, not inflation-adjusted GDP or quantities sold.

That sector split complicates a uniform recovery story. It does not establish which workers received higher wages or which firms added jobs. A shared reporting period does not make industrial profits and household income interchangeable.

CATL’s unaudited third-quarter filing supplies a company example for January–September 2026. Profit attributable to listed-company shareholders rose 31.60 percent year on year. Consolidated cash paid to and on behalf of employees was RMB 32.04 billion, against RMB 23.41 billion in the same 2025 period. These amounts convert the filing’s thousand-yuan units to billions; they are not individual earnings.

The cash-flow line includes payments on employees’ behalf. Without matched average staffing, hours, payment components and worker-level distribution, its increase cannot establish higher pay per worker. CATL’s consolidated group is also a different population from NBS’s domestic household and industrial surveys. Matching periods makes the question sharper; it does not establish that profits caused wage gains.

The half-year filing narrows the accounting question: its employee-benefit note separates wages, bonuses, allowances and subsidies from welfare and contributions. These are January–June accrual and liability movements, not the January–September cash measure above. A decrease in amounts payable cannot be labeled verified take-home pay. The inspected notes supply no matched hours or individual earnings.

The remaining link is within the workplace: period-matched payroll, hours, staffing and commission arrangements, checked against worker accounts. The household release now supplies a national income distribution measure, and the filing supplies company payments. Neither tells us whether Lv or a CATL worker earned more for the same work.

The causal claim that weak consumption necessarily forces the observed export growth is not established by these releases.