Summary

This report presents the findings of the third annual study of public perceptions of air quality in Almaty. The study was prepared by the Almaty Air Initiative Foundation in collaboration with political scientist Serik Beissembayev. At the Foundation’s request, PwC Kazakhstan reviewed the report on a pro bono basis and provided its expert opinion.

The purpose of the study is to assess residents’ levels of awareness, willingness to act, and actual behavior in relation to air pollution, and to track changes compared with the 2024 and 2025 survey waves. The Index is designed as a management tool: a reproducible annual indicator that can be used to assess the effectiveness of the city’s environmental and information policies and to compare Almaty with other cities.

For the first time in the current sociological study, the Air Quality Awareness Index was calculated as a composite indicator that brings survey results onto a common scale from 0 to 100.

The Index assumes that awareness is not limited to knowledge but consists of a sequence of four stages. Residents first perceive air pollution as a threat, then learn about its causes and ways to check air quality, then develop a willingness to act, and only after that translate this willingness into actual behavior. Each stage is measured separately, allowing the Index to show not only the overall level of awareness, but also the specific point at which the chain may break down. A higher Index score indicates a more comprehensive understanding of the problem and more active behavior.

The fieldwork was conducted from 18 February to 11 March 2026. Total of 1,000 Almaty residents aged 18 and above were surveyed. The sample is representative of the city’s adult population by gender, age, and district of residence and covers all eight administrative districts. The statistical margin of error does not exceed ±3.1 percentage points. The Index methodology is based on the approaches set out in the Handbook on Constructing Composite Indicators (OECD / Joint Research Centre of the European Commission).

The overall Index score for Almaty is 41 out of 100, corresponding to the category “low level, predominantly passive attitude.”

Behind this figure lies a stable pattern that constitutes the central finding of the study. Air pollution is recognized as a problem almost universally across the city: the overwhelming majority of residents consider air pollution dangerous to health, and this proportion has increased for the third consecutive year. However, knowledge of where to check air quality and what measures are implemented remains limited to a minority of residents. Willingness to act declines noticeably as soon as a measure affects residents’ own everyday behavior. Actual practices are the least widespread, and they are dominated by strategies of personal protection and avoidance rather than participation.

In other words, concern about air pollution is high, while the knowledge and practical experience required to translate this concern into action remain insufficient. The gap between emotional perception of the problem and practical behavior is the most pronounced characteristic of the overall profile, and data from the three survey waves indicate that it has not yet narrowed.

One finding that extends beyond the environmental agenda deserves particular attention. The proportion of residents seriously considering moving away from Almaty has increased noticeably over the past two years, and most of them cite air quality as the main reason. At the same time, this intention remains reversible: half of potential movers would abandon their plans to relocate if the situation improved significantly. Air quality is therefore identified by the study not only as an environmental factor, but also as a demographic one.

Methodology

The study is based on a sociological survey of Almaty residents. 

The fieldwork was conducted from 18 February to 11 March 2026, with a sample of 1,000 respondents; the sample is representative of the city’s adult population (aged 18 and above) by gender, age, and district of residence and covers all eight administrative districts of the city. The statistical margin of error does not exceed ±3.1 percentage points at a 95% confidence level; the margin of error is higher for subsamples such as individual districts and age groups.  

The profile consists of four sub-indices measuring different dimensions of awareness: affective (emotional perception of the threat), cognitive (knowledge and awareness), conative (attitudes and willingness to act), active — actual practices.  

The methodology follows the approaches set out in the Handbook on Constructing Composite Indicators (OECD / Joint Research Centre of the European Commission; Nardo et al., 2005).

The Index measures the social perception of risk and does not replace environmental monitoring data.  

Key Findings 

The overall Index score and profile stand at 41.2 out of 100, corresponding to the category “low level, predominantly passive attitude”.  

The profile shows a consistent decline in engagement as the progression moves from emotions to action: high concern (70.4) does not translate into knowledge (37.6), willingness to act (40.2), or actual practices (16.8).  

The gap between concern and action is 53.5 points, making it the defining feature of the current situation. 

Key observations: 

1. 81% of respondents consider the state of air quality dangerous to health, up from 74%, while 63% report a high level of concern about the issue.
2. Despite the availability of air quality monitoring tools, awareness of them remains limited. Only 20% of respondents know where they can check air quality, while 62% have not heard of any relevant platforms. At the same time, official air quality data are used by only around 5% of respondents.
3. Support for different measures declines substantially as the personal costs associated with them increase. Systemic measures whose implementation falls primarily on the government, or the energy sector receive relatively high levels of support: development of public transport – 54%, conversion of CHP plants to gas – 52%, and restrictions on older vehicles – 51%. By contrast, measures involving additional costs or restrictions for residents themselves receive noticeably less support, for example, 28% of respondents support paid entry to the city center for all vehicles.
4. The most pronounced gap is observed between concern about the problem and practical action. Only 12% of residents regularly check air quality data, almost half of respondents (46%) have taken no environmental action over the past year, and 17% use no measures to protect themselves from polluted air. Existing practices are more strongly oriented towards adaptation to current conditions, while participation in collective initiatives, civic oversight, and support for environmental projects occurs substantially less often than respondents’ stated willingness to engage in such activities.
5. Awareness, risk perception, and behavioral patterns vary considerably across population groups, which may limit the effectiveness of uniform communication approaches. The chain “concern → knowledge → action” manifests differently across groups: younger respondents demonstrate higher levels of awareness but lower engagement in practical action, while older respondents are more likely to express concern but less likely to use available information sources. The most balanced indicators are observed among respondents aged 25–34. The gap in knowledge of information channels between groups with different levels of education is nearly eightfold (7% versus 56%). Having children is also associated with a higher perception of the threat posed by air pollution, regardless of respondents’ other characteristics (51% versus 42%).
   

Conclusion 

The study identifies a persistent pattern: a high level of concern about air pollution coexists with limited awareness of available information sources and a comparatively low level of practical engagement.

The findings provide a clearer understanding of residents’ attitudes towards air pollution and may serve as an additional source of evidence for the development of communication and environmental initiatives.

Section 2. Air Quality Awareness Index: Overall Picture 

2.1. How the Index is structured 

The Air Quality Awareness Index aggregates four dimensions of residents’ attitudes towards air pollution, arranged in a logical sequence: affective (emotional perception of the threat) → cognitive (knowledge and awareness) → conative (attitudes and willingness to act) → active (actual practices). Each dimension is assessed on a scale from 0 to 100, while the overall Index is calculated as the average of the four sub-indices (see Section 8 for methodology). 

The overall Index score is interpreted according to the following scale: 0–25 — extremely low level, 26–50 — low level, predominantly passive attitude, 51–75 — pronounced concern with limited activity, 76–100 — high level of awareness and active behavior. 

2.2. Overall Index Score 

The overall Air Quality Awareness Index score for Almaty is 41.2 out of 100.  

This corresponds to the category “low level, predominantly passive attitude.” The city is positioned near the upper boundary of this range but does not reach the “pronounced concern with limited activity” category (51–75). Fuller picture is provided by the four sub-indices shown below. 

 

The profile shows a broadly declining pattern along the chain “threat perception → knowledge → willingness → action.” 

The following changes are observed at each transition between dimensions: 

• from emotional perception of the threat (affect) to knowledge and awareness (cognition): a decline of 32.7 points. Emotional recognition of the problem is not supported by sufficient knowledge, most concerned residents do not know where to check air quality: 62% do not know where to access such information (Section 4), while 56% are unaware of any initiatives currently being implemented (Section 4). 

• from knowledge and awareness (cognition) to willingness to act (conation): an increase of 2.6 points. Existing knowledge is translated into stated attitudes and intentions; attitudes are therefore not the main bottleneck in the profile. 

• from willingness to act to actual action: a decline of 23.4 points, from 40.2 to 16.8. Stated willingness does not translate into actual practices: 46.4% took no environmental action over the past year, while 16.8% use no measures to protect themselves from polluted air (Section 6). 

The gap between the affective and active dimensions is 53.5 points, which is the defining characteristic of the current profile. In other words, the high level of concern about air pollution (70.3) is not accompanied by a comparable level of practical action among residents (16.8).

2.3. Interpretation of the Profile 

The structure of the profile supports three main conclusions, which are examined in greater detail in the following sections.

First, the lack of awareness is driven not by a lack of interest in the issue, but by a shortage of practical information. The affective dimension is the only component within the medium-to-high range: most residents perceive air pollution as a serious threat to health (81.1% consider it dangerous or very dangerous), while 57.2% of respondents considering relocation cite air quality as the main reason for this decision. The problem is therefore already perceived as significant and potentially dangerous. At the same time, a high level of concern is not always accompanied by sufficient knowledge of the causes of pollution, available data sources, and possible ways to respond (Section 8).

Second, the main constraints are concentrated in the cognitive and active dimensions. The low cognitive sub-index score is largely driven by limited awareness of how to check air quality (indicator B1 — 29.1) and of existing initiatives (B3 — 29.9). Knowledge of pollution sources is more developed (B2 — 53.8), although the findings indicate certain distortions in how these sources are perceived (Section 4). The low active sub-index score (16.8) reflects the infrequent practice of regularly checking air quality data (D1 – 14.4), limited use of protective measures (D2 – 23.7), and low engagement in environmental action (D3 – 12.2).

Third, a high level of concern coexists with limited knowledge and practical action. The gap between the affective and active dimensions is 53.5 points: the affective sub-index reaches 70.3, while the active sub-index stands at only 16.8. This indicates a substantial gap between perceptions of the air pollution problem and the actual level of engagement in actions aimed at addressing it. Where access to clear and relevant information sources is limited, perceptions of the problem may be shaped to a greater extent by personal experience and informal communication channels.

2.4. Trends: Comparison with the 2024–2025 Surveys 

The current survey is the first in which the Air Quality Awareness Index has been calculated. However, changes in public perceptions can be assessed using comparable individual indicators: 

 Taken together, the indicators show differing trends across the components. Cognitive indicators are improving gradually, albeit from a very low base: knowledge of air quality monitoring channels has doubled over two years but still reaches only one fifth of the population. Affective indicators have stabilized at a high level, while behavioral indicators show no comparable improvement. The gap between awareness and action should therefore be regarded as not yet narrowing, and its magnitude may serve as a target indicator for assessing the effectiveness of the city’s information and environmental policies. 

Sections 3–6 examine each of the four dimensions in sequence, covering the sub-index score, the distribution of responses across the respective indicators, and differences between socio-demographic and territorial groups. Section 7 brings these differences together into group profiles. 

Section 3. Affective Dimension: Emotional Perception of the Problem 

3.1. Sub-index score and indicator structure 

Affective sub-index measures population’s emotional response to air pollution: assessment of air quality in the area of residence, a degree of personal concern, and perception of health risks. 

The score of 70.3 is the highest among the four sub-indices and the only one falling within the medium-to-high range. Clear gradient is observed within the sub-index: perception of health risks (A3) produces the highest score, while assessment of air quality in one’s own area (A1) is more moderate. Residents perceive the overall threat as more severe than the situation in their immediate surroundings.

3.2. Assessment of air quality (indicator A1) 

According to the current survey, 30.6% of respondents rate air quality in their area of residence as “poor” and 18.8% as “very poor.” Overall, 49.4% provide a negative assessment. Neutral assessment (“satisfactory”) is given by 31.8%, while 17.9% provide a positive assessment (“good” — 15.5%, “very good” — 2.4%).  

Share of negative assessments increased from 42% in 2024 to 50% in 2025 and has stabilized at approximately this level in the current wave. This increase may reflect not so much an actual deterioration in air quality as growing awareness: as residents become better informed about the issue (Section 4), they may begin to assess familiar conditions more critically.  

Marked differences are observed across districts. Negative assessments dominate (≥50%) in the Almaly (64.6%), Zhetysu (54.2%), Auezov (52.9%), and Bostandyk (50.0%) districts. The most favorable perceptions are found in Nauryzbay District, where 65.7% rate air quality as satisfactory or better, while the share giving the most negative assessment, “very poor,” is the lowest at 8.8%. The highest proportion of extremely negative assessments is recorded in Zhetysu District, where 31.3% describe air quality as “very poor,” approximately twice the citywide level. 

3.3. Personal concern (indicator A2) 

Air pollution concerns 31.3% of residents “very strongly” and 32.2% “strongly,” bringing the combined level of pronounced concern to 63.5%. Neutral position is reported by 22.6%, while 13.3% are little or not at all concerned.  

The overall level of pronounced concern remains unchanged from the summer 2025 wave (63%), indicating that the affective background has stabilized at a high level following the increase observed between the 2024 and 2025 waves. 

The age pattern of concern is non-linear. The highest levels of “very strong” concern are recorded among economically active middle-age groups: 39.1% among those aged 35–44 and 35.5% among those aged 25–34. The lowest level is among respondents aged 18–24 (15.1%), more than twice as low as among middle-aged respondents. The 65+ group is also below the overall average (23.9%). Thus, the issue is felt most acutely by the working-age population, which also experiences the greatest symptom burden (Section 6). 

Differences are also observed across Almaty districts. Residents of Bostandyk and Zhetysu districts report the highest levels of concern: 43.4% and 33.7%, respectively, selected “very strongly concerned”. By contrast, only 28.4% of residents in Nauryzbay District and 26.3% in Turksib District reported the same level of concern. 

Strong concern increases with education level: among residents with secondary education or below, 27.4% report strong concern; among those with higher education, the figure is 34.1%; and among holders of academic degrees, it reaches 44.4%, the highest level. 

Having children also substantially increases concern: among parents of children under 18, 41.3% are “very strongly” concerned about air pollution, compared with only 24.8% among respondents without children.  

A consistent pattern is observed across the different socio-demographic dimensions: concern increases with income, education, and having children. In other words, the more resources an individual has, the more seriously they tend to perceive the problem. This has two implications that become apparent in later sections. On the one hand, better-resourced groups are more active than others in protecting themselves from pollution (Section 6). On the other hand, they are also more likely to consider moving away (Section 3.5).

Comparison of the territorial patterns for A1 and A2 reveals a geographical mismatch: the district with the highest concern, Bostandyk, is not among those with the worst perceived air quality, whereas Turksib and Nauryzbay, despite substantial actual exposure associated with private housing and industrial zones, display the lowest levels of concern. The affective response is therefore shaped not only by actual pollution levels but also by the socio-economic characteristics of district residents. More affluent districts tend to express greater concern even where environmental conditions are comparable or better, while districts facing objectively higher exposure report lower levels of concern. 

3.4. Perception of health risks (indicator A3) 

Risk perception is the highest-scoring component of the affective dimension. Air pollution is considered “very dangerous” to health by 45.6% of residents and “rather dangerous” by another 35.5%, (total 81.1%), a further 11.5% give an intermediate assessment, describing pollution as “dangerous in some respects but not in others,” while 6.4% do not perceive a significant threat (“rather not dangerous” — 4.8%, “not dangerous at all” — 1.6%). Compared with the 2025 wave, the combined share rating air pollution as “dangerous/very dangerous” increased from 74% to 81.1%. This is the only affective indicator that continues to rise while the others have stabilized. 

• Having children increases perceived risk: among parents of minors, 51.1% consider air pollution “very dangerous,” compared with 41.8% among respondents without children.  

• By age, the highest shares of “very dangerous” assessments are found among respondents aged 35–44 (51.1%) and 55–64 (50.8%), while the lowest are among those aged 18–24 (36.3%) and 65+ (35.1%). 

The parental perspective therefore strengthens risk perception beyond age differences and may provide one of the most universal frames for public communication. 

3.5. Behavioral consequences of emotional perception: migration intentions 

Within the structure of the report, migration intentions are treated as an extreme manifestation of the affective dimension: intention to leave the city represents the strongest form of response to a perceived threat and, unlike indicators of the active dimension, does not require everyday behavioral practices. 

One in four residents is seriously considering leaving Almaty: 21.0% think about relocation frequently and discuss it with their family, 2.8% have a firm decision and 2.1% are in the process of moving. Combined, these groups account for 25.9% of respondents. A further 32.4% have considered moving but have no specific plans, while only 41.3% have never considered leaving. In the 2024 survey wave, 17% were seriously considering relocation. 

Air pollution is the most frequently cited reason for potential relocation: 57.2% of those considering moving name it as the main reason, far ahead of low income and limited career prospects (22.8%) and the desire to provide a better future for their children (20.6%). At the same time, these intentions are largely reversible: among potential migrants, 50.1% would abandon plans to relocate if air quality improved significantly, whereas for 48.0% such an improvement would not be decisive. As for preferred destinations, countries outside the near region account for the largest share (26.7%), followed by other cities in Kazakhstan (24.2%). 

Among respondents not concerned at all about air pollution, 61.7% have never considered leaving Almaty, while only 16.7% seriously consider relocation: 11.7% think about it and 5.0% have a firm decision or are in the process of moving, among those strongly concerned, 39.0% seriously consider relocation: 31.6% think about it frequently, 4.2% have firm decision, and 3.2% are in the process of moving. The relationship between the share of potential migrants and the level of concern provides direct quantitative evidence of the link between the affective dimension and migration intentions. 

Socio-demographic differentiation: 

• Education. The higher the level of education, the more frequently respondents consider relocation: among holders of master’s or academic degrees, 40.7% either frequently consider leaving, have already decided to leave, or are preparing to relocate. This is twice the share among respondents with secondary education (20.2%). 

• Age. Relocation is considered most frequently by respondents aged 25–34 — 40.2% (based on the share of the three responses indicating  intention to leave), the lowest share is recorded among respondents aged 65+ (10.4%). 

• Children. Among parents with one or two children, 28–30% consider relocation, compared with 23% among respondents without children.  

Seasonal migration is widespread among those who are generally open to leaving, including 54.5% would like to leave the city during the winter if they had the opportunity, and 15.0% already do so every year, this amounts to 69.5%, while for 30.5% this is not typical. 

 The highest share of residents who either wish to leave the city for the winter or already do so every year is found among those most concerned about air pollution: 19.0% and 61.9%, respectively. The highest share of residents who neither leave nor plan to leave for the winter, 60.9%, is recorded among those who are not concerned at all about air pollution in Almaty. 

3.6. Conclusions on the affective dimension 

The affective sub-index (70.3) is the only component of the Index within the medium-to-high range and is what keeps the overall score near the upper boundary of the “low” category. The affective background has stabilized: following the increase between the 2024 and 2025 waves, the shares of negative assessments and pronounced concern have remained unchanged, while only the perception of health risks continues to rise (74% → 81.1%). Concern is concentrated among working-age middle-aged groups and is further strengthened by parenthood.  

Migration intentions are directly linked to affect and are largely reversible: more than half of potential migrants would remain in the city if air quality improved, making air quality not only an environmental factor but also a demographic factor in the city’s long-term sustainability. 

The affective dimension is the highest-scoring component of the profile. Section 4 examines the second link in the chain: the extent to which strong threat perception is accompanied by knowledge of the causes of pollution, ways to check air quality, and initiatives currently being implemented. 

Section 4. Cognitive Dimension: Knowledge and Awareness 

4.1. Sub-index score and indicator structure 

The cognitive sub-index measures public awareness: knowledge of where and how to check air quality, perceptions of the main sources of pollution, and awareness of initiatives currently being implemented. 

The cognitive sub-index is separated from the affective sub-index by the largest gap in the profile — 32.7 points. The internal structure of this dimension is uneven: relatively well-developed perceptions of pollution sources coexist with limited knowledge of monitoring tools and existing initiatives. Residents have a general understanding of causes of the problem but have limited access to data on its current state and to information about measures being implemented. 

4.2. Knowledge of air quality monitoring channels (indicator B1) 

20.4% of respondents confidently know where to check air quality, 17.4% have heard something about available sources but are unsure and 62.2% do not know.  

Awareness has steadily improved: the share who know where to check air quality rose from 10% in 2024 to 20.4%, while the uninformed share fell from 74% to 62.2%.  Nevertheless, the baseline remains low: four out of five residents still do not have confident access to air quality data, even though monitoring tools, including public maps, mobile applications, and AAI’s independent sensor network, have become substantially more accessible over the same period. In other words, the problem is not the absence of tools, but the fact that people are unaware of them. 

The trend across survey waves represents the most consistent positive development observed in the study: 

Knowledge of monitoring channels is stratified by age and education. 

• By age, it is highest among respondents aged 25–34 (30.8%) and lowest among those aged 65+ (6.7%) and young people aged 18–24 (19.2%). 

• The education gap is even wider: awareness rises from 7.1% among those with secondary education or below to 55.6% among those with master’s or academic degrees. The difference between the two extremes is nearly eightfold. 

Those who check air quality use specialized mobile applications (37.3%), follow websites (27.0%). Social media and forums also play a notable role (16.7%), as do household air quality sensors (15.7%), while data provided by city authorities are used by only 4.9%.  

Notably, even within this relatively informed group, 32.7% still rely on their own senses, particularly smell and the visible presence of smog.

The age-related digital divide is a structural feature of the cognitive dimension: older age groups combine high levels of concern with minimal access to data. The part of the population most affectively engaged with the issue is therefore cognitively disconnected from objective information. 

4.3. Knowledge of initiatives to improve air quality (indicator B3) 

Only 15.9% can name a specific air quality initiative,  27.6% have heard something and 55.8% have heard of none.  

Compared with the 2025 survey, the share of respondents with confident knowledge increased from 10% to 15.9%, while the share who had not heard of any initiatives declined from 61% to 55.8%, this is comparable to the increase in awareness of monitoring channels but still remains low. 

The most frequently cited initiatives relate to gasification and CHP conversion to gas (26.4%), followed by transport regulation (15.7%), solid fuel restrictions in the private sector (13.2%), environmental initiatives (10.1%), urban greening (8.8%), and restrictions on older vehicles (8.2%). Taza Qazaqstan campaign is mentioned by 5.0% of respondents.

Comparing spontaneous awareness with the list of measures actually implemented shows that public awareness of many existing initiatives remains limited. Several major infrastructure programs are mentioned by only a relatively small share of respondents. At the same time, issues that have received broader public and media attention, such as LEZ, demonstrate higher levels of recognition.

4.4. Perceptions of the sources of pollution 

Residents identify road transport as the main source of air pollution, cited by 79.7% of respondents. This is followed by CHP plants (51.8%) and uncoordinated urban development that obstructs air circulation (47.3%). Less frequently mentioned sources include tree cutting (35.4%), waste burning (31.9%), industrial enterprises (31.8%), coal-based heating in the private housing sector (29.8%), construction activity (23.5%), and natural factors (23.5%). Compared with the 2025 wave, attribution to transport remained broadly stable (82% → 79.7%), while the share identifying CHP plants increased (45% → 51.8%), possibly influenced by public discussion of the role of CHP plants and their conversion to gas. 

Perceptions of pollution sources vary by age. CHP plants are cited most often by respondents aged 25–34 (72.0%) and 35–44 (63.9%), while private-sector heating is mentioned mainly by younger groups (42%) and rarely by those aged 65+ (12.7%). Attribution to transport remains consistently high across all age groups, at approximately 80–83%. 

These findings are important for policy, as public support for measures is shaped by how residents perceive the structure of pollution sources. Measures targeting sources that are underestimated by the public, such as restrictions on coal burning in the private housing sector, may therefore initially receive lower levels of public support. 

4.5. Trust in official information as a factor in the cognitive deficit 

Alongside limited awareness of available information and current initiatives, the study finds relatively low confidence in the city authorities’ ability to improve air quality in the coming years. Overall, 53.3% of residents do not believe that such an improvement will occur (25.5% — “do not believe at all,” 27.8% — “rather do not believe”), while 26.5% take the opposite view (5.6% — “fully believe,” 20.9% — “rather believe”). Further 18.1% were unable to provide a clear assessment. Distrust is highest among respondents aged 25–64, where it ranges from 53% to 57%. It is somewhat lower among young people aged 18–24 (49.3%) and respondents aged 65+ (44.8%).

At the same time, official air quality data are used by only 4.9% of respondents who check relevant information, the lowest share among all channels considered.

Section 5 examines the third link in the chain: what attitudes and stated willingness to act emerge at the observed levels of concern and awareness.

4.6. Conclusions on the cognitive dimension 

• The largest gap between sub-indices is observed between the affective (70.3) and cognitive (37.6) dimensions and amounts to 32.7 points. This means that the level of concern about air pollution substantially exceeds the level of awareness of the issue.
• The lowest scores are recorded for knowledge of monitoring tools (B1 — 29.1) and initiatives (B3 — 29.9), while perceptions of pollution sources are more developed (B2 — 53.8). Both indicators show sustained improvement while remaining at a relatively low level. 
• Awareness differs substantially across population groups: by education level, the gap is nearly eightfold (7.1% versus 55.6%). By age, an inverse relationship is observed between concern and awareness: older age groups show higher levels of concern while having lower access to information. 
• Perceptions of pollution sources are skewed: transport remains overestimated as a source of pollution, while the contribution of the private housing sector is underestimated. 
• The study also records a combination of low confidence in the ability of the authorities to address air pollution (53.3%) and limited use of official information sources.

Section 5. Conative Dimension: Attitudes and Willingness to Act 

5.1. Sub-index score and indicator structure 

The conative sub-index measures the attitudinal component of awareness: stated support for systemic measures and respondents declared personal willingness to take action and accept restrictions. 

 

Transition from cognitive to conative dimension is not accompanied by a decline: the attitudinal sub-index (40.2) is 2.6 points higher than the cognitive sub-index. Existing knowledge is translated relatively fully into stated attitudes. Meaningful internal gap is observed within the dimension: support for systemic measures (44.2) is approximately 1.2 times higher than personal willingness to act (36.1). On average, respondents support around 4 of the 9 proposed measures but are personally willing to take part in only 1.8 of 5 actions. 

5.2. Support for systemic measures (indicator C1) 

Support for individual measures in the current wave is distributed as follows: 

Structure of support forms three tiers. The upper tier (46–54%) consists of infrastructure and regulatory measures whose costs are borne primarily by the government, the energy sector, or a limited group of violators: public transport, conversion of CHP plants to gas, restrictions on older vehicles, urban greening, and construction controls. The middle tier (38–41%) includes measures with more specific target groups: school buses, control of waste burning, and gasification of the private housing sector. The lower tier consists of a single measure that directly restricts the everyday behavior: paid entry to the city center (28.0%), which trails the next closest measure by 10 percentage points. The composition of the upper tier — transport, energy, green spaces, and urban development — is reproduced across all survey waves and may be regarded as a stable area of public consensus. 

5.3. Personal willingness to act 

Distribution of stated willingness in the current wave is as follows: 

Total of 13.1% state that they are willing to undertake all of the listed actions, while 14.0% are unwilling to take any of them. The pattern reflects the same logic as support for systemic measures, but even more clearly: the only action with willingness above 50% is participation in collective environmental activities, which requires neither financial expenditure nor changes to everyday habits. All actions involving financial costs, such as paying more for transport (20.3%) or making donations (13.0%), giving up habitual behavior, such as not driving into the city center (16.0%), or making regular efforts in civic oversight, such as filing complaints (17.5%), fall within a narrow range of 13–20%. 

Willingness varies by age. Willingness to participate in environmental activities increases with age, from 38.4% among respondents aged 18–24 to 65.3% among those aged 55–64, reflecting the established tradition of community clean-up activities. By contrast, willingness to pay more for environmentally friendly transport declines with age, from 31.5% among younger respondents to 11–13% among older age groups. The share of respondents unwilling to take any action is highest at both ends of the age distribution: 19.9% among those aged 18–24 and 20.9% among those aged 65+. 

5.4 Conclusions on the conative dimension 

• The conative sub-index (40.2) is higher than the cognitive sub-index; attitudes are therefore not the main bottleneck in the profile. 

• At the same time, within the dimension, support for systemic measures (44.2) exceeds personal willingness (36.1) by approximately 1.2 times, while for individual actions the gap can be as large as twofold: support falls by half when the costs of a measure are transferred directly to the respondent. 

• The areas of stable public consensus (transport, energy, urban greening, and development) are reproduced across survey waves. 

Section 6 examines the final link in the chain: actual practices. The overlap between certain items in the lists of stated willingness and actions actually taken allows declarations to be compared directly with behavior (Section 6.4).

Section 6. Active Dimension: Actual Practices 

6.1. Sub-index score and indicator structure 

The active sub-index measures actual behavior: the frequency of checking air quality data, the use of personal protection measures, and environmental actions taken over the past year. 

Active sub-index is the lowest component of the profile: it is 24.4 points below the overall Index and 53.5 points below the affective sub-index and is the only component in the “extremely low” range. All three indicators fall within the 12–24 point range, indicating a generally low level of behavioral engagement. 

6.2. Use of air quality data 

Total of 68.7% of residents never check air quality data in applications or on websites. Another 8.7% check it occasionally, (once in six months) 9.8% several times a month; 5.9% several times a week; and 6.4% almost every day. Thus, regular checking, defined as several times a week or more, is characteristic of only 12.3% of residents, while more than two thirds never check air quality data at all. 

Regular use declines sharply with age: the share of those who never check air quality rises from 52.1% among respondents aged 18–24 to 85.8% among those aged 65+. Respondents aged 25–34 are the most active users of air quality data, with regular checking reaching 18.7%. 

Since a stable habit of checking air quality is limited by awareness of available channels (20.4%, Section 4.2), while only 12.3% actually check the data regularly, the gap between knowing about a channel and using it consistently represents a separate opportunity for improving the active dimension. 

6.3. Personal protection measures (indicator D2) 

Use of protection measures in the current wave is distributed as follows: 

 The pattern is skewed towards avoidance and compensation strategies: two most common responses, leaving the city (43.5%) and taking vitamins (38.9%), do not reduce exposure at the place of residence and do not involve control over the quality of the air being inhaled. Instrumental measures that directly reduce exposure are used far less frequently: air purifiers (23.1%), window filters (11.5%), and masks (9.5%). Compared with the 2025 wave, the share of air purifier users has remained unchanged (23% → 23.1%), indicating stagnation despite a twofold increase in awareness. 

The use of air purifiers varies sharply by age: it is highest among respondents aged 25–34 (35.5%) and falls to 10.5% among those aged 55–64. In other words, older groups with the highest levels of concern (Section 3.3) make the least use of technical protection. Share of those taking no protective measures is relatively even (13–22%) and is highest among respondents aged 45–54 (21.7%). 

6.4. Environmental actions over the past year (Indicator D3) 

Over the past year, 44.6% of residents planted trees, (in the yard or as part of clean-up activities); 5.1% participated in public hearings or environmental campaigns; 4.9% filed an official complaint about a pollution source; and 3.7% donated money to environmental organizations. At the same time, 46.4% took none of these actions. 

Comparing actual actions with stated willingness (Section 5.3) for matching items provides a direct measure of the attitude–behavior gap: 

The conative–active gap is most pronounced in forms of participation that require interaction with institutions and organizations. These are also the practices with the lowest stated willingness to participate. Stated willingness is translated into action unevenly: simple collective activities are carried out at nearly the stated level, whereas participation in civic oversight and financial support for initiatives occurs almost three times less often than respondents say they are willing to engage in such actions. Age analysis also points to low youth engagement: among respondents aged 18–24, 56.2% took no action at all, the highest share across all age groups.

6.5. Context of the active dimension 

Behavioral practices take place against a background of widespread direct experience of air pollution. According to the previous survey wave, 80% of residents reported negative effects of air pollution on their health, most commonly headaches and fatigue (40%), allergic reactions (39%), irritation of mucous membranes (36%), and breathing problems (35%). The greatest burden was reported among the economically active 25–44 age group. Almost half of residents reported having health conditions that increase vulnerability to polluted air, particularly chronic respiratory conditions (21%) and cardiovascular conditions (12%). 

Comparison with the current wave shows that the 25–44 age group, which bears the greatest symptom burden, also demonstrates the highest level of concern (Section 3.3) and the highest level of protective activity. For most residents, however, reported symptoms do not translate into either instrumental protection or regular use of air quality data. 

6.6. Conclusions on the active dimension 

1. The active sub-index (16.8) is the lowest component of the profile, with low scores consistently observed across all three indicators (12–24 points). 
2. Protective practices are biased toward evasion and compensation with limited use of instrumental measures; the proportion of purifier users does not increase between waves despite a twofold increase in awareness. 
3. Only 12.3% of residents regularly check air quality data, while 46.4% took no environmental action over the past year. 
4. The gap between willingness and action is selective: low-cost collective practices are implemented almost entirely, while institutionally mediated ones are implemented almost three times less frequently than stated.  
5. Reducing barriers to instrumental protection and developing a habit of regularly checking air quality data are the main opportunities for improving the active sub-index. 
   

Section 7. Social and Territorial Differentiation of Awareness

Sections 3–6 identified socio-demographic differences within individual indicators. This section brings them together into group profiles and identifies the point in the chain “concern → knowledge → willingness → action” at which the breakdown occurs for different groups.

7.1. Age gradient: inverse relationship between knowledge and concern

Age is the most pronounced and comprehensively measured axis of differentiation in the current wave. Knowledge is concentrated among younger and middle-age groups, concern among middle-aged and older groups, while only respondents aged 25–34 display a balanced profile.

 The chain breaks at different points. Young people aged 18–24 is moderately informed but affectively detached, (the lowest level of strong concern – 15.1%) and passive (the highest level of inactivity – 56.2%). The gap therefore occurs at the “knowledge → concern/action” stage. The 25–34 group is the only one with a balanced profile: the highest awareness of channels (30.8%), the highest rate of regular checking (18.7%), and use of air purifiers (35.5%). Older groups (55+) combine high concern and strong perceptions of risk with minimal access to data (6.7–15.3% know where to check air quality, up to 85.8% never check it) – the breakdown occurs at the “concern → knowledge” stage.

7.2. Education, family status, and geography

Education strengthens the cognitive dimension: knowledge of monitoring channels increases from 7.1% among those with secondary education or below to 55.6% among holders of academic degrees. This nearly eightfold gap is the steepest socio-demographic gradient identified in the study.

Parenthood strengthens the affective dimension: having children under 18 increases the share of respondents who consider air pollution “very dangerous” from 41.8% to 51.1%, the broad consensus around the risks posed to children makes the parental perspective one of the most universal drivers of awareness.

Gender differences are observed primarily in the affective dimension, as recorded in previous waves. In the current wave, the main gradients are defined by age and education.

7.3. Territorial differentiation: mismatch between exposure and concern

Perception patterns vary across districts and do not fully overlap, indicating that territorial social factors matter alongside actual exposure.

 The most negative self-assessments of air quality are concentrated in Almaly, Zhetysu, Auezov, and Bostandyk districts. Migration pressure is highest in Almaly (31.9%) and Bostandyk (30.9%). The most acute cognitive gap is in the Alatau district: with an average level of danger perception (50% “very dangerous”), only 9% of residents here know where to test the air – three to four times lower than in Bostandyk (34%) or Almaly (26%).

Bostandyk District, by contrast, combines the highest perception of risk (59%) with the highest awareness of channels (34%) – an “engaged exposure” profile, in which concern is translated into information-seeking activity. These contrasts help define territorial priorities for information policy: districts with high concern but low levels of knowledge, particularly Alatau, require more proactive and targeted information support.

7.4. Typical awareness profiles

Taken together, the gradients can be grouped into four analytical profiles, distinguished by the point at which the chain breaks:

7.5. Conclusions on differentiation

1. Awareness is stratified primarily by age and education. 
2. Knowledge is concentrated among younger and middle-age groups and among the highly educated, with an eightfold education gap. Concern is concentrated among working-age middle-aged groups, while only the 25–34 age group demonstrates a balanced profile.
3. Parenthood is the most universal driver of awareness. 
4. Territorial perception patterns do not fully overlap. Cognitive gaps in individual districts, particularly Alatau, do not directly correspond to levels of concern and require targeted information policies. 

Sections 3–7 present the findings of the current survey wave. Section 8 describes the conceptual model of the Index, the indicator structure, calculation algorithm, data requirements, and limitations on the interpretation of results.

8. Conceptual Model of the Index

The Air Quality Awareness Index is a composite social index designed to quantify how the population perceives the problem of air pollution and how this perception is translated into attitudes, intentions, and behavior. The Index does not measure actual pollution levels and does not replace environmental monitoring data.

Awareness is defined as a construct comprising four dimensions: 

• affective dimension – emotional response and subjective perception of threat;
• cognitive dimension – knowledge and awareness of pollution sources, monitoring methods, and initiatives being implemented;
• conative dimension – attitudes, intentions, and willingness to change;
• active dimension – actual practices and behavioral responses.

The model draws on risk perception theory, health behavior models, and the Knowledge–Attitude–Behavior (KAB/KAP) approach and assumes a sequential but not strictly linear process: “threat perception → knowledge → attitudes → action”: different configurations are possible, including high levels of knowledge combined with low activity, or strong concern accompanied by limited support for reforms – identifying such profiles is one of the objectives of the Index (Section 7).

The empirical basis for the Index included an analysis of international practices for measuring environmental awareness and comparable composite indices, the results of AAI surveys conducted in 2024–2025, three focus group discussions with Almaty residents, and a pilot online survey (n = 120) used to test question wording and scales. The methodology follows the approaches set out in the Handbook on Constructing Composite Indicators (OECD / Joint Research Centre of the European Commission).

The Index was developed by the Almaty Air Initiative Foundation in collaboration with political scientist Serik Beissembayev.

8.1. Indicator structure

The Index is calculated using 11 indicators:

8.2. Calculation algorithm

The calculation is carried out in four stages.

Stage 1. Coding Each response option is assigned a score. A higher score corresponds to a higher level of awareness, concern, or activity. Uncertain responses, such as “difficult to answer,” are treated as missing values and excluded separately for each indicator.

Stage 2. Normalization. Indicators are converted to a 0–100 scale:

Normalized score = (Actual score − Minimum score) / (Maximum score − Minimum score) × 100

Example: a value of 4 on a 1–5 scale gives: (4 − 1) / (5 − 1) × 100 = 75

Stage 3. Sub-indices. Each of the four sub-indices is calculated as the arithmetic mean of the normalized values of the indicators included in that dimension.

Stage 4. Overall Index. Awareness Index = (S_aff + S_cog + S_con + S_act) / 4 Equal weighting reflects the conceptual equivalence of the four dimensions and corresponds to the baseline methodological scenario.

Interpretation of the overall score:

These categories are normative and may be adapted depending on the objectives of the analysis.

8.3. Data requirements and status of calculations for the current wave

The methodology requires the Index to be calculated using individual anonymized respondent-level data provided by the survey contractor in a machine-readable format (SPSS .sav or Excel .xlsx), together with variable coding and descriptions of the sample and fieldwork procedures. Providing only aggregated tables or presentation materials is insufficient and does not allow the calculation to be reproduced.

The accuracy of preliminary estimates is ±3–5 points. Conclusions regarding the Index category and the structure of the profile remain robust under the documented assumptions. Until recalculation is completed, numerical values may only be published if clearly labelled as preliminary.

8.4. Limitations

The Awareness Index measures the social perception of environmental risk and cannot be used as a substitute for environmental or public health indicators. The results reflect subjective assessments, which may differ from monitoring data.

The indicators are sensitive to the informational and institutional context. Media agendas, high-profile events, and public campaigns may have short-term effects, particularly on the affective dimension. Interpretation of annual dynamics must take into account the context of the field period.

The Index is also subject to the standard limitations of survey research, including socially desirable responses, which are most likely in the conative dimension where willingness is self-reported, question wording effects, and differences in how respondents interpret scales. The results should be interpreted as indicators of trends and comparative differences rather than as a precise measure of individual behavior.

Changes in question wording and response options between survey waves reduce the comparability of individual time series. For long-term tracking, wording should therefore be standardized and kept consistent across waves.

8.5. Ethical principles

Data collection is conducted in accordance with generally accepted ethical standards for social research, including voluntary participation, informed consent, the right to decline to answer any question, and the absence of pressure or manipulative wording. Particular attention is given to questions concerning health risks. Their wording should not unnecessarily increase anxiety.

All data are anonymized. No direct personal identifiers are collected, access to raw datasets is restricted, and results are published exclusively in aggregated form that does not allow individual respondents or households to be identified.

The Index results are not used: to stigmatize social groups or territories (including when publishing regional differences – sections 3, 7); as a basis for sanctions or repressive measures; or in manipulative or misinformative communications. The purpose of the Index is to support policy analysis, improve the quality of public dialogue, and contribute to the development of targeted communication strategies. Responsible use requires transparent presentation of the methodology, clear acknowledgement of limitations, and avoidance of oversimplified or politically biased interpretations. In this report, this requirement is operationalized through the documentation of methodological assumptions.