雅思阅读 066 — Remote Work at Five: 82 Studies Say Hybrid Wins

改编自 OsakaWire（2026-09，Florent Herisson），基于对 82 项研究、3,574 个估计值的元分析。

🎧 课文朗读音频（约 1 分半）

Reading Passage

In March 2026, 22.6 per cent of American workers teleworked — 24.9 per cent of employed women and 20.5 per cent of employed men. Working from home has settled at roughly a quarter of all paid workdays in the United States, about 3.5 times its 2019 level. The return-to-office campaign of 2023 to 2026 did not reverse the shift; it renegotiated it. What remains unsettled is what five years of randomised and quasi-experimental evidence say the arrangement did to output, to careers, and to the buildings it emptied.

The stabilisation is the least disputed finding in the field. The Current Population Survey has held the national telework rate inside a band of 17.9 per cent to 23.8 per cent since October 2022. Measured in days rather than people, full days worked at home account for about a quarter of all paid workdays, roughly 3.5 times the 2019 share.

The arrangement that actually settled is hybrid, not fully remote. Fully remote work covers about a tenth of employees. Australia's Productivity Commission puts fully remote roles at about 5 per cent of positions and roles requiring five days on site at about 6 per cent — both are tails rather than the distribution. The typical white-collar job in 2026 is neither: it is two or three days in an office and the remainder at a kitchen table.

The average employee will accept a pay cut of about 7 to 8 per cent for the option to work from home two or three days a week. The physical saving is 72 minutes of commuting a day, about 40 per cent of which is returned to the employer as extra work.

The two experiments that disagree

The same intervention measures as a 13 per cent performance increase in one randomised trial and an 18 per cent productivity decline in another. Both are gold-standard experiments with clean assignment and large effects. Both results are correct.

In 2010, economists randomised call-centre staff at the Chinese travel agency Ctrip into nine months of home working. Performance rose 13 per cent: nine points from working more minutes per shift, and four points from more calls per minute in a quieter environment. Attrition halved and reported satisfaction rose.

In 2017, the same design was applied to 235 data-entry workers in Chennai, India. Those assigned home were 18 per cent less productive. Most of the gap was visible on the first day — an environment effect rather than an adaptation effect — and the rest accumulated through slower learning across two months. Workers who said they preferred home working were 27 per cent less productive when given it.

Two variables separate the trials, and neither is the policy. The first is the counterfactual workspace: Ctrip's home workers escaped an open-plan call floor, while Chennai's left a purpose-built office for households that offered no equivalent quiet. The second is the learning content of the task: answering airline queries is a stable routine performed by experienced staff, while data entry at a new employer is a skill acquired on the job.

Remote work does not have a single productivity coefficient. It has a coefficient conditional on the task and the room.

Where the compromise landed

The meta-analysis pools 3,574 estimates from 82 studies and finds that the effect varies by sex, parental status and global region as well as by arrangement. Hybrid work — one or two days a week at home — produces no measurable loss in performance and, in many cases, higher retention and satisfaction. Fully remote work, by contrast, runs roughly 10 per cent behind on average, with the largest penalties concentrated on junior staff in learning-intensive roles.

The literature also identifies a "career tax": employees who work fully from home are promoted at lower rates than comparable peers who come into the office, largely because informal visibility and sponsorship — not formal output — drive advancement. Junior employees who never enter an office lose access to the apprenticeship-style learning that comes from overhearing senior conversations.

The practical conclusion five years on is modest. For routine, individually measurable work, remote delivery is close to free. For collaborative, learning-intensive work, periodic in-person contact appears necessary. The "hybrid settlement" — two or three days in the office, the rest at home — is not a compromise between two extremes but a distinct arrangement that most evidence suggests works.

Questions

Questions 1–5: TRUE / FALSE / NOT GIVEN

1. The US telework rate has changed dramatically since October 2022.
2. The majority of American white-collar workers now work fully from home.
3. The average employee would take a pay cut of 7–8% to work from home two or three days a week.
4. The Ctrip and Chennai experiments used different research designs.
5. The Ctrip trial found that home workers were 13% more productive.

Questions 6–10: Choose the correct letter, A, B, C or D.

6. In the Chennai experiment, most of the productivity gap

   • A. appeared on the first day
   • B. emerged only after six months
   • C. was caused by lower motivation
   • D. disappeared with training

7. The key difference between the Ctrip and Chennai trials was

   • A. the country where the study took place
   • B. the age of the workers
   • C. the task type and the home environment
   • D. the pay scale

8. According to the meta-analysis, fully remote work on average

   • A. improves productivity by 13%
   • B. performs roughly 10% behind office-based work
   • C. has no effect on output
   • D. reduces costs by half

9. The "career tax" refers to

   • A. taxes paid on remote income
   • B. lower promotion rates for fully remote employees
   • C. office rental costs
   • D. commuting expenses

10. The passage concludes that the most evidence-supported arrangement is

    • A. five days in the office
    • B. fully remote work
    • C. hybrid work with two or three days in the office
    • D. no fixed arrangement

Questions 11–13: Complete the sentences. Choose NO MORE THAN TWO WORDS.

11. The meta-analysis pools 3,574 estimates from __________ studies.
12. Home workers in the Ctrip study saved roughly __________ minutes of commuting each day.
13. Junior remote employees lose access to __________-style learning from senior colleagues.

Answers

1. FALSE (基本稳定在 17.9-23.8%)
2. FALSE (完全远程只占约 10%)
3. TRUE
4. FALSE (都是随机对照试验，设计相同)
5. TRUE
6. A
7. C
8. B
9. B
10. C
11. 82
12. 72
13. apprenticeship

Glossary

• telework /ˈtelɪwɜːk/ n. 远程办公
• hybrid /ˈhaɪbrɪd/ adj. 混合的
• meta-analysis /ˌmetə əˈnæləsɪs/ n. 元分析
• attrition /əˈtrɪʃn/ n. 人员流失
• coefficient /ˌkəʊɪˈfɪʃnt/ n. 系数
• quasi-experimental /ˌkwɑːzi ɪkˌsperɪˈmentl/ adj. 准实验的
• apprenticeship /əˈprentɪʃɪp/ n. 学徒制
• retention /rɪˈtenʃn/ n. 留存
• productivity /ˌprɒdʌkˈtɪvəti/ n. 生产率