Oct 07, 2021
Hong Kong Public Opinion Research Institute Press Conference – Press Materials
Detailed Findings
Speakers:
Kim-Wah Chung – Deputy CEO, HKPORI
Kenneth Chan – Associate Professor, Department of Government & International Studies, HKBU; Director, Comparative Governance & Policy Research Centre; Member, Election Observation Project
Edward Tai – Manager (Data Science), HKPOR
Oct 7, 2021
Hong Kong Public Opinion Research Institute Press Conference – Press Materials
Special Announcement
The predecessor of Hong Kong Public Opinion Program (HKPOP) was The Public Opinion Programme at The University of Hong Kong (HKUPOP). “POP” in this release can refer to HKPOP or its predecessor HKUPOP.
Abstract
After Chief Executive Carrie Lam delivered the Policy Address yesterday, POP conducted an instant survey on the same day and released part of the findings last night. Apart from random landline and mobile numbers, this survey also included samples from our “Hong Kong People Representative Panel” (i.e., a panel comprising randomly recruited samples) within “HKPOP Panel”, interviewed by telephone or invited through email to complete an online survey. Our telephone survey began at around 1:30pm till around 9:30pm, while our online survey started at around 1:30pm and ended at around 8pm yesterday. A total of 936 successful cases were collected, including 228 random landline samples, 260 random mobile samples, 127 panel telephone survey samples and 321 panel online survey samples. The raw data have been weighted by population statistics and proportions of different sampling frames to ensure data representativeness.
Our survey shows that after excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year. As for CE Carrie Lam, her latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered. Moreover, after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative. The instant survey describes people’s instant reaction toward the Policy Address. Their reactions later remain to be seen.
The effective response rate of the survey excluding panel samples is 54.8%. The maximum sampling error of percentages is +/-4%, that of net values is +/-7% and that of ratings is +/-2.7 at 95% confidence level.
Contact Information
| Date of survey | : | 6/10/2021 |
| Survey method | : | (1a) Random landline telephone survey (1b) Random mobile telephone survey (2a) Telephone survey targeting “Hong Kong People Representative Panel” within “HKPOP Panel” (2b) Online survey with email invitation targeting “Hong Kong People Representative Panel” within “HKPOP Panel” |
| Target population | : | Cantonese-speaking Hong Kong residents aged 18 or above |
| Sample size[1] | : | 936 (including 228 random landline samples, 260 random mobile samples, 127 panel telephone survey samples and 321 panel online survey samples) |
| Effective response rate | : | 54.8% (excluding panel samples) |
| Sampling error[2] | : | Sampling error of percentages not more than +/-4%, that of net values not more than +/-7% and that of ratings not more than +/-2.7 at 95% conf. level |
| Weighting method | : | The raw data comes from 4 different sampling frames. It is rim-weighted by two sets of weighting factors simultaneously. The first set of weighting factors comprises population figures provided by the Census and Statistics Department, they include (a) the gender-age distribution of the Hong Kong population from “Mid-year population for 2020”, (b) educational attainment (highest level attended) distribution from “Women and Men in Hong Kong – Key Statistics (2020 Edition)”, and (c) economic activity status distribution from the last source. The second set of weighting factors is adjusted based on the relative target sample size of different sub-sampling frames, namely, random telephone survey using landline numbers set at 5 units, random telephone survey using mobile numbers set at 5 units, telephone survey of randomly pre-selected panel members set at 6 units, and online survey of randomly pre-selected panel members set at 4 units. |
[1] This figure is the total sample size of the survey. Some questions may only involve a subsample, the size of which can be found in the tables below.
[2] All error figures in this release are calculated at 95% confidence level. “95% confidence level” means that if we were to repeat a certain survey 100 times with different random samples, we would expect 95 times having the population parameter within the respective error margins calculated. Because of sampling errors, when quoting percentages, journalists should refrain from reporting decimal places, whereas one decimal place can be used when quoting rating figures.
Latest Figures
People’s satisfaction figures with this year’s Policy Address are summarized below together with the previous findings:
| Date of survey | Sample size[3] | Appraisal of Policy Address | |||||
| Satisfaction rate[4] | Half-half | Dissatisfaction rate[4] | Net satisfaction rate | Mean value[4] | Rating of Policy Address | ||
| 6/10/21 | 621 | 25+/-4%[6] | 13+/-3%[6] | 50+/-4%[6] | -25+/-7%[6] | 2.4+/-0.1[6] | 34.2+/-2.7[6] |
| 25/11/20 | 512 | 19% | 9% | 64% | -46% | 2.0 | 27.2 |
| 16/10/19 | 679 | 17%[6] | 8%[6] | 65%[6] | -47%[6] | 2.0[6] | 29.7[6] |
| 10/10/18 | 534 | 33%[6] | 24% | 34%[6] | -1%[6] | 2.9[6] | 48.5[6] |
| 11/10/17 | 526 | 48%[6] | 28%[6] | 14%[6] | 34%[6] | 3.5[6] | 62.4[6] |
| 18/1/17 | 512 | 34%[6] | 22% | 29%[6] | 5%[6] | 3.0[6] | 52.3[6] |
| 13/1/16 | 522 | 19%[6] | 23% | 39% | -20%[6] | 2.5[6] | 41.1[6] |
| 14/1/15 | 503 | 30%[6] | 24%[6] | 35% | -5%[6] | 2.8 | 49.5[6] |
| 15/1/14 | 611 | 36% | 30%[6] | 31%[6] | 5% | 3.0 | 54.1[6] |
| 16/1/13 | 759 | 36%[6] | 35% | 24%[6] | 11%[6] | 3.1 | 56.4[6] |
| 12/10/11 | 816 | 47%[6] | 32% | 18% | 28%[6] | 3.3 | 59.1 |
| 13/10/10 | 747 | 41%[6] | 33%[6] | 19%[6] | 22%[6] | 3.2 | 58.9[6] |
| 14/10/09 | 462 | 30% | 37% | 28% | 2% | 3.0 | 53.5 |
| 15/10/08 | 515 | 31%[6] | 35%[6] | 26%[6] | 4%[6] | 3.0 | 53.8[6] |
| 10/10/07 | 602 | 52%[6] | 29%[6] | 10%[6] | 42%[6] | 3.5 | 65.2[6] |
| 11/10/06 | 445 | 30%[6] | 37% | 22%[6] | 8%[6] | 3.0 | 55.8[6] |
| 12/10/05 | 377 | 48%[6] | 33% | 9%[6] | 39%[6] | 3.5 | 66.4[6] |
| 12/1/05 | 391 | 38%[6] | 30% | 20%[6] | 18%[6] | 3.2 | 56.3[6] |
| 7/1/04 | 381 | 25% | 26% | 33%[6] | -8% | 2.8 | 49.3 |
| 8/1/03[5] | 377 | 22%[6] | 29% | 27% | -5% | 2.8 | 51.6[6] |
| 10/10/01 | 433 | 29% | 33% | 28% | 1% | 3.0 | 56.7 |
| 11/10/00 | 262 | 25%[6] | 28% | 31% | -6%[6] | 2.9 | 55.2 |
| 6/10/99 | 236 | 31%[6] | 30% | 25%[6] | 6%[6] | 3.0 | 57.3 |
| 7/10/98 | 508 | 22%[6] | 35%[6] | 35%[6] | -14%[6] | 2.8 | — |
| 8/10/97 | 534 | 45% | 30%[6] | 14%[6] | 31% | 3.4 | — |
[3] Respondents who did not answer this question because they had not heard of / did not have any knowledge of the Policy Address have been excluded. Before March 2020, weighted count was used to report subsample size. Starting from March 2020, raw count was used instead.
[4] Collapsed from a 5-point scale. The mean value is calculated by quantifying all individual responses into 1, 2, 3, 4, 5 marks according to their degree of positive level, where 1 is the lowest and 5 the highest, and then calculate the sample mean.
[5] The 2003 Policy Address instant survey was conducted for two days. Only figures registered in the first day of fieldwork are listed in this table for direct comparison and analysis.
[6] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
After excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. The mean score is 2.4, meaning between “quite dissatisfied” and “half-half” in general. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year.
Figures on various Chief Executives’ popularity before and after the Policy Address Speech since 1997 are summarized as follows:
| Popularity of Tung Chee-hwa | ||||||||
| Date of PA Speech | 8/10/97 | 7/10/98 | 6/10/99 | 11/10/00 | 10/10/01 | 8/1/03 | 7/1/04 | 12/1/05 |
| Rating before the PA | 65.8 | 55.8 | 54.0 | 48.2 | 48.4 | 46.6 | 42.9 | 47.2 |
| Rating at instant survey | 66.1 | 56.1 | 54.3 | 50.7 | 50.6 | 47.3 | 44.6 | 48.4 |
| Change in rating | +0.3 | +0.3 | +0.3 | +2.5[8] | +2.2[8] | +0.7 | +1.7[8] | +1.2 |
| Popularity of Donald Tsang | ||||||||
| Date of Policy Address Speech | 12/10/05 | 11/10/06 | 10/10/07 | 15/10/08 | 14/10/09 | 13/10/10 | 12/10/11 | |
| Rating before the PA | 68.0 | 62.9 | 65.8 | 52.7 | 55.2 | 55.4 | 48.4 | |
| Rating at instant survey | 67.4 | 59.8 | 64.4 | 53.9 | 54.2 | 56.2 | 50.6 | |
| Change in rating | -0.6 | -3.1[8] | -1.4[8] | +1.2 | -1.0 | +0.8 | +2.2[8] | |
| Net approval rate before the PA | 68% | 48% | 48% | 5% | 7% | -1% | -45% | |
| Net approval rate at instant survey | 65% | 36% | 48% | 10% | 8% | 0% | -41% | |
| Change in net approval rate[7] | -3% | -12%[8] | — | +5% | +1% | +1% | +4% | |
| Popularity of CY Leung | ||||||||
| Date of Policy Address Speech | 16/1/13 | 15/1/14 | 14/1/15 | 13/1/16 | 18/1/17 | |||
| Rating before the PA | 48.9 | 45.6 | 40.6 | 37.5 | 41.3 | |||
| Rating at instant survey | 52.2 | 48.9 | 44.8 | 37.0 | 41.7 | |||
| Change in rating | +3.3[8] | +3.3[8] | +4.2[8] | -0.5 | +0.4 | |||
| Net approval rate before the PA | -20% | -31% | -39% | -44% | -44% | |||
| Net approval rate at instant survey | -11% | -24% | -35% | -54% | -57% | |||
| Change in net approval rate[7] | +9%[8] | +7%[8] | +4% | -10%[8] | -13%[8] | |||
| Popularity of Carrie Lam | ||||||||
| Date of Policy Address Speech | 11/10/17 | 10/10/18 | 16/10/19 | 25/11/20 | 6/10/21 | |||
| Rating before the PA | 59.6 | 52.3 | 22.3 | 30.8 | 33.9+/-2.0 | |||
| Rating at instant survey | 61.1 | 47.6 | 22.7 | 26.8 | 30.5+/-2.2 | |||
| Change in rating | +1.5 | -4.7[8] | +0.3 | -4.1[8] | -3.4[8] | |||
| Net approval rate before the PA | 10% | 4% | -65% | -48% | -46+/-5% | |||
| Net approval rate at instant survey | 23% | -10% | -64% | -57% | -48+/-5% | |||
| Change in net approval rate[7] | +13%[8] | -14%[8] | +1% | -9%[8] | -2% | |||
[7] Instant surveys on Policy Address included CE’s approval rate since 2004, so it is not listed under Tung’s series.
[8] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
Recent figures on Carrie Lam’s popularity before and after the Policy Address speech are as follows:
| Date of survey | 19-22/7/21 | 9-12/8/21 | 20-26/8/21 | 6-10/9/21 | 16-23/9/21 | 6/10/21 | Latest change |
| Sample size | 1,000 | 1,002 | 1,003 | 1,000 | 1,036 | 936 | — |
| Response rate | 48.5% | 49.4% | 52.9% | 44.2% | 44.1% | 54.8% | — |
| Latest findings | Finding | Finding | Finding | Finding | Finding | Finding & error | — |
| Rating of CE Carrie Lam | 34.7 | 35.1 | 33.8 | 35.6 | 33.9 | 30.5+/-2.2 | -3.4[9] |
| Vote of confidence in CE Carrie Lam | 20% | 20% | 20% | 24%[9] | 20%[9] | 19+/-3% | -1% |
| Vote of no confidence in CE Carrie Lam | 68% | 66% | 68% | 65% | 66% | 67+/-3% | +1% |
| Net approval rate | -48% | -46% | -48% | -41% | -46% | -48+/-5% | -2% |
[9] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
Instant survey shows that CE Carrie Lam’s latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered.
The survey also gauged the change of people’s confidence in the future of Hong Kong after CE Carrie Lam delivered her Policy Address. Results are as follows:
| Date of survey | Sample size[10] | Confidence in the future of Hong Kong | |||
| Increased | Unchanged | Decreased | Net effect on confidence | ||
| 6/10/21 | 621 | 25+/-4%[11] | 19+/-3% | 50+/-4%[11] | -24+/-7%[11] |
| 25/11/20 | 512 | 17%[11] | 16%[11] | 63% | -46% |
| 16/10/19 | 679 | 12%[11] | 22% | 61%[11] | -49%[11] |
| 10/10/18 | 534 | 23%[11] | 25%[11] [12] | 45%[11] | -22%[11] |
| 11/10/17 | 526 | 40%[11] | 39% | 19%[11] | 21%[11] |
| 18/1/17 | 511 | 24%[11] | 36%[11] | 32%[11] | -8%[11] |
| 13/1/16 | 521 | 16%[11] | 31% | 44%[11] | -27%[11] |
| 14/1/15 | 501 | 22% | 35% | 38%[11] | -16% |
| 15/1/14 | 846 | 24%[11] | 38% | 32%[11] | -9%[11] |
| 16/1/13 | 913 | 31% | 38%[11] | 23% | 8% |
| 12/10/11 | 957 | 29% | 45% | 21% | 8%[11] |
| 13/10/10 | 914 | 31%[11] | 45% | 18%[11] | 14%[11] |
| 14/10/09 | 749 | 27%[11] | 47%[11] | 22%[11] | 5%[11] |
| 15/10/08 | 761 | 23%[11] | 38%[11] | 32%[11] | -9%[11] |
| 10/10/07 | 388 | 53%[11] | 31%[11] | 7%[11] | 46%[11] |
| 11/10/06 | 431 | 25%[11] | 51%[11] | 16%[11] | 9%[11] |
| 12/10/05 | 476 | 54%[11] | 33%[11] | 5%[11] | 49%[11] |
| 12/1/05 | 658 | 34% | 41% | 12%[11] | 22%[11] |
| 7/1/04 | 602 | 32%[11] | 40% | 16%[11] | 16%[11] |
| 8/1/03[13] | 513 | 25% | 40%[11] | 21% | 3% |
| 10/10/01 | 591 | 22% | 50%[11] | 21%[11] | 1%[11] |
| 11/10/00 | 292 | 22%[11] | 40% | 15% | 7%[11] |
| 6/10/99 | 233 | 40%[11] | 36%[11] | 16%[11] | 24%[11] |
| 7/10/98 | 505 | 21% | 52% | 22% | -1% |
[10] Respondents who did not answer this question because they had not heard of / did not have any knowledge of the Policy Address have been excluded. Before March 2020, weighted count was used to report subsample size. Starting from March 2020, raw count was used instead.
[11] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
[12] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level because of a change in the weighting method. If the previous weighting method was used, the changes would not have gone beyond the sampling errors.
[13] The 2003 Policy Address instant survey was conducted for two days. Only figures registered in the first day of fieldwork are listed in this table for direct comparison and analysis.
Results show that after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative.
Data Analysis
Our latest Policy Address instant survey shows that after excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year.
As for CE Carrie Lam, her latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered.
Moreover, after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative.
The instant survey describes people’s instant reaction toward the Policy Address. Their reactions later remain to be seen.