JOINT EUROPEAN COMMISSION – OECD WORKSHOP ON INTERNATIONAL DEVELOPMENT OF BUSINESS AND CONSUMER TENDENCY SURVEYS
BRUSSELS
14 – 15 NOVEMBER 2005
Consumer Confidence Index for Turkey Ece Oral
Central Bank of the Republic of Turkey Research and Monetary Policy Department
- November 2005 -
Consumer Confidence Index for Turkey Ece Oral *
Central Bank of the Republic of Turkey Research and Monetary Policy Department
[email protected]
Abstract The aim of this study is to give brief knowledge about the Consumer Tendency Survey (CTS) carried out with cooperation of the Central Bank of the Republic of Turkey (CBRT) and the State Institute of Statistics (SIS). The second aim is to construct a consumer confidence index by using the questions of the CTS. The purpose of constructing the consumer confidence index is to find out consumer tendencies and expectations for general economic course. The most appropriate index is formed and its performance is tested in this study.
*
The views expressed in this paper are those of the author and do not necessarily correspond to the views of the Central Bank of the Republic of Turkey.
1. Introduction Consumer behavior plays an important role in providing economic decision-makers and economic forecasters with necessary information about future expectations. Optimistic consumer confidence may cause desire for making large expenses and increase the tendency for borrowing, while pessimism may cause consumers to reduce their expenditures, to review their financial situations. Consumer surveys provide regular assessments of consumer attitudes and expectations and are used to evaluate economic trends and prospects. The surveys are designed to explore why changes in consumer expectations occur and how these changes influence consumer spending and saving decisions. They also provide the necessary data for the measurement of the consumer confidence (Kershoff, 2000). 2. Consumer Tendency Survey and the Purpose The State Institute of Statistics (SIS) and the Central Bank of Turkey (CBRT) have jointly launched a study within the framework of “Consumer Survey” Protocol in 2003. The Consumer Tendency Survey (CTS), which is the source for the consumer confidence index, was annexed to Household Labour Force Survey in the form of a module. The pilot study of the survey was carried out with the aim of testing the questionnaire design and method implemented between April 2003 and December 2003. During the pilot-test period, a number of modifications were made in the questionnaire design and some methods were tested in sample-selection. Following an eight-month pilot-test period, it was decided to announce the survey results on a monthly basis. CTS has been constructed in order to find out consumer tendencies and expectations for general economic course, job opportunities, personal financial standing and market developments in order to assess their expenditure behavior as well as their expectations, and therefore deciding monthly consumer tendencies in the short-run. CTS covers four fields in measuring consumer tendencies and expectations that are Personal Financial Standing, General Economic Situation, Expenditures and Price Expectations. 3. Scope of Survey The target population used for the survey is the individuals at the age of 15 and above having a job in rural and urban areas of Turkey that provide income. Therefore, the survey covers all individuals at the age of 15 and above who are employed as samples in Household Labour Force Survey meeting these criteria. The frame is updated four times a year and the date of the last update/ revision is in February 2005. 2
The survey includes all settlements (rural and urban) in Turkey. Settlements with a population of 20 001 and over are defined as urban while settlements with a population of 20 000 or less are defined as rural. The number of participants in the frame is approximately 7000. Regular employee, casual employee, employer and self-employed constitute 53, 10, 6 and 31 percent of the status in employment groups in the survey, while they constitute 57, 10, 0 and 33 percent in the whole population respectively. Similarly, the participants aged 15 to 24, 25 to 34, 35 to 54, 55 and more constitute 14, 34, 43 and 9 percent of the age groups in the survey, while they constitute 22, 30, 38 and 10 percent in the whole population respectively. Therefore, it can be said that status in employment and age groups represent the population. However, in income category, the population coverage cannot be attained. 4. Sampling Method 2000 individuals at the age of 15 and above having income from an economic activity were selected from Household Labour Force Survey in December 2003, January and February 2004 to represent Turkey in general on the basis of age, income and status in employment groups. Whole individuals at the age of 15 and above having income from an economic activity have been interviewed at house as samples in Household Labour Force Survey since March 2004. The average number of units included in the survey is approximately 7000. For the module, age, status in employment and income groups have been taken into consideration in addition to general weight of Household Labour Force Survey. Monthly Household Labour Force Survey is carried out via a rotation system, in which 50% new respondents and 50% former respondents being interviewed for the second time after three months. The State Institute of Statistics' field interviewers conduct computer-based, face-to-face interviews (CAPI) with individuals, who have been selected on basis of the above-mentioned criteria under this rotation system. The fieldwork starts on Monday of the first week of the month and lasts 2 weeks. The length of the reference period to which the survey data relates, is one month. The average delay between the end of the reference period and the date of public release of data is 22 days. 5. Data Characteristics The first step is to convert the numbers of answers to each of the three/four/five reply options into percentages. The percentage of positive replies (much more better, a little bit better, increase sharply, increase slightly, very likely, fairly likely, it is right time now), unchanged replies (remain the same, neither the right time nor the wrong time) and negative replies (much more worser, a little bit worser, fall sharply, fall slightly, not likely, not at all likely, it 3
is not the right time now) indicate the direction of change of a variable (question). The net balance is then calculated as the difference between the percentages of positive and negative responses. Question 7 has nominal scale. All the other questions are presented in a 3, 4 or 5-point Likert scale format. The questions of the CTS are given in Appendix. Neither an imputation technique nor a seasonal adjustment procedure is used. 6. Weights Used in the Survey The revised Household Labour Force Survey results are used in order to weigh the Household Labour Force Survey. The weights used for the Household Labour Force Survey are design weights, non-response adjustments, external control weights and final adjustment factor. For Consumer Tendency Survey, age, status in employment and income groups are taken into consideration in addition to general weight of Household Labour Force Survey. No additional weighting is used in constructing the consumer confidence index. 7. Calculation Method Indices are compiled in accordance with the balance method of European Union. The balance is calculated as the difference between the percentages of positive and negative responses and 100 is added to this difference, thus forming a separate diffusion index for each question. Then, the general index is calculated by taking arithmetic means of diffusion indices of the questions included in consumer confidence index. The index is evaluated between 0 and 200. If it is above 100, it means consumer confidence is optimistic. If it is below 100, consumer confidence is pessimistic. 100 refers to neutral opinion in consumer confidence. There is no base period for confidence indicator calculated in index form. 8. Dissemination of the Survey Results The results of the previous month are announced following the first working day of the 22nd of
each
month
through
web
sites
of
CBRT
and
SIS,
http://www.tcmb.gov.tr/tuketanket/consumereng.html and http://www.die.gov.tr respectively. 9. Consumer Confidence Index Confidence indicators derived from business and consumer survey results are of key importance in assessing short-term economic developments. These sentiment indicators give
4
crucial information on business and consumers assessments of the current economic situation and their intentions and expectations for the future (Nilsson, 2000). Consumer confidence measures were formulated in the late 1940's by George Katona at the University of Michigan as a way to include empirical measures of consumer expectations into the models of spending and saving behavior. Katona summarized his view by saying that consumer spending depends on both their “ability and willingness to buy.” Katona meant discretionary purchases; the income and assets of consumers; consumers’ assessments of their future job and income prospects by using the words "spending", "ability" and "willingness" respectively. When consumers become optimistic they increase their spending, and when they become pessimistic they decrease their spending and increase their precautionary saving. The term "Discretionary purchases" is used for the things that people buy that are not the necessities of life. There is no reason to consider that the consumer confidence should play a role in determining the timing of necessary expenditures. Homes, vehicles, furniture, appliances, home electronics, PCs, and other large household durables can be considered as discretionary purchases. These are relatively infrequent and large purchases, often involve the use of credit, and are typically planned purchases—just the type of purchases that would critically depend on consumers’ expectations about their future economic situation (Curtin, 2002). For constructing a consumer confidence index for Turkey, the survey questions providing qualitative information on the current situation as well as on the expectations for the next months are used. CTS has been conducted since December 2003 which means the number of observations is insufficient to perform statistical analysis like cross-correlation or peak-trough analysis. Therefore, the variables used in constructing confidence index of EU are investigated and most of them are used in order to construct the confidence index for the Turkish case. The variables related to the purchasing power of the consumer (at present and over the next six months) that can be thought as a proxy for the income of consumers (questions 1 and 2); the variable about job opportunities (next six months) (question 5); the general economic situation (next three months) (question 4); buying time condition of durable goods (present) (question 7), which can be considered as a leading indicator for the demand for durable goods are chosen in order to construct consumer confidence index. The constructed confidence index can be thought as a leading indicator by including three expectations questions. The methodology used for the construction of the confidence indicator in this study is the same as the one derived in EU (User Guide, 2003). The calculation and the
5
interpretation of the index are described in Section 7 above. The diffusion indices of survey questions and the Consumer Confidence Index are given in Table 4. 10. Performance of the Consumer Confidence Index The questions that are used to construct the Consumer Confidence Index and the constructed confidence index are compared with the conventional quantitative realizations to evaluate their degree of consistence with the realizations. The Consumer Confidence Index’s performance is tested in the following way. The Consumer Confidence Index is compared with the year-to-year changes of industrial production index in Figure 1 for illustrating its performance. Figure 1 shows that the index seems to behave parallel to the IPI. Figure 1 Industrial Production Index and Consumer Confidence Index 25
115 Consumer Confidence Index
20
110 IPI (year to year change %) (right scale)
15
105
10 100
5
Jul.05
Jun.05
Ma.05
Apr.05
Marc.05
Feb.05
Jan.05
Dec.04
Nov.04
Oct.04
Sep.04
Aug.04
Jul.04
Jun.04
Ma.04
Apr.04
Marc.04
-5 Feb.04
90 Jan.04
0 Dec.03
95
The Consumer Confidence Index is also compared with year-to-year changes of GNP, GDP, private consumption expenditures and durable goods expenditures at fixed (1987) prices in Table 1. It can be seen that the diffusion indices of questions 1 and 7, Consumer Confidence Index, private consumption expenditures and durable goods expenditures illustrate similar pattern. Table 1
Period
Private Durable Consumption Goods Expenditures* Expenditures* (%) (%)
GDP* (%)
GNP* (%)
Diffusion Diffusion Index of Index of Question 1** Question 7**
Consumer Confidence Index**
2004Q1
12.4
48.0
11.8
13.9
100.0
134.5
111.4
2004Q2
18.4
61.4
14.4
15.7
97.3
138.6
108.3
2004Q3
7.3
28.9
5.3
5.7
93.6
135.0
103.1
2004Q4
3.6
-5.7
6.3
6.6
93.4
134.1
103.9
2005Q1
4.0
1.5
4.8
5.3
91.6
132.9
104.2
2005Q2
4.4
2.6
4.2
3.4
89.8
132.7
99.9
* Year-to-year changes at fixed (1987) prices ** Three-month averages 6
The consumer, housing and automobile loans are investigated and year-to-year changes of realized loans are given in Table 2 for the survey period. The consumer, automobile and other loans got the highest value in March 2004 and then started to fall down. The diffusion index of question 7 also demonstrates this downward trend (Table 4). It can also be seen from Figure 2 that the amounts of sales of automobiles and white goods show parallel tendency with the automobile and consumer loans respectively. Table 2
Consumer Loans
Housing Automobile Individual Loans Loans Other Loans Credit Cards Year- to- year changes (%) 60.5 206.0 111.4 35.6 69.2 228.7 122.8 42.4 97.8 258.6 135.4 42.9 137.4 308.8 150.3 47.4 197.8 383.9 181.5 58.5 246.8 423.5 190.4 51.7 252.1 422.7 177.3 58.6 253.1 402.1 165.1 69.8 248.6 355.1 149.0 60.5 246.3 285.0 132.4 61.9 213.9 190.0 108.1 77.2 195.6 135.9 94.0 77.2 177.2 83.6 80.4 91.3 168.7 60.1 62.3 75.2 153.9 51.8 83.0 72.9 147.7 40.2 76.2 69.6 148.1 27.7 64.3 61.4 143.7 17.0 54.7 57.5 165.3 16.5 56.0 51.1 207.6 19.4 58.4 43.3 225.1 21.9 57.3 39.8
125.9 139.7 159.1 187.0 233.8 257.7 249.6 236.9 216.1 190.9 148.3 122.6 95.5 76.4 82.8 75.3 66.1 57.6 62.3 72.6 77.0
Dec.03 Jan.04 Feb.04 Marc.04 Apr.04 Ma.04 Jun.04 Jul.04 Aug.04 Sep.04 Oct.04 Nov.04 Dec.04 Jan.05 Feb.05 Marc.05 Apr.05 Ma.05 Jun.05 Jul.05 Aug.05
Figure 2 The Amount of Sales of Automobiles and White goods 600000
60000
500000
50000
400000
40000
300000
30000
200000
Durable Goods
20000
100000
Automobiles (right scale)
10000
7
Jul.05
Jun.05
Ma.05
Apr.05
Marc.05
Feb.05
Jan.05
Dec.04
Nov.04
Oct.04
Sep.04
Aug.04
Jul.04
Jun.04
Ma.04
Apr.04
Marc.04
Feb.04
Jan.04
0 Dec.03
0
Finally, the diffusion index of question 5 and the unemployment rates are examined over the survey period. The unemployment rate has decreased in the second quarter of 2004 and after that it started to increase until the first quarter of 2005. Table 3
Period
Labour force participation rate (%)
Unemployment Rate (%)
Non-agriculture unemployment rate (%)
2003Q4
47.1
10.3
14.6
2004Q1
45.9
12.4
17.3
2004Q2
49.2
9.3
13.6
2004Q3
50.6
9.5
14
2004Q4
48.4
10
14.1
2005Q1
46.8
11.7
15.4
2005Q2
49.3
9.2
12.5
The moving averages of the diffusion index of question 5 and the 6-month changes in unemployment rate are calculated in order to test the consistency with the realizations in Figure 3. The question 5 (job opportunities expectations) shows an opposite structure with the changes in unemployment rate as expected. Figure 3
Unemployment Rate and Question 5
98.0
0.30
96.0
0.20
94.0
0.10
92.0 0.00 90.0 88.0
Question 5 (monthly average)
86.0
Change in Unemployment Rate (%) (right scale)
-0.10 -0.20
8
2005Q2
2005Q1
2004Q4
-0.30 2004Q3
84.0
YEAR & Month 2003 12 2004 1 2 3 4 5 6 7 8 9 10 11 12 2005 1 2 3 4 5 6 7 8
Consumer confidence index
110.3 111.4 111.9 111 111 107.3 106.6 105.4 101.2 102.8 103.7 102.7 105.2 105.4 105.2 102.1 100.4 100.3 99.1 99.2 97.5
Purchasing power (present)
97.9 100.8 99.5 99.8 99.2 95.8 96.9 94.6 92.7 93.4 94.4 93.0 92.7 92.8 92.3 89.8 89.6 89.6 90.1 90.3 89.6
Purchasing power (next 6 months)
107.5 108.3 106.5 106.1 104.7 100.2 101.9 97.9 94.7 96.1 96.7 96.0 97.5 99.2 98.7 95.8 94.2 94.8 93.8 92.8 91.3
General economic situation (present) 114.8 115.4 114.1 112.4 113 106.2 106.1 103.3 97.9 100.5 101.5 100.7 102.8 106.5 103.7 98.0 95.0 93.6 92.5 92.4 91.1
General economic situation (next 3 months) 118.3 117.6 115.1 113.1 114.1 106.1 107.5 105 98.9 102.8 103.5 100.8 104.7 108.6 105.5 100.5 97.1 95.9 93.6 94.2 93.0
Job opportunities (next 6 months) 107.5 104.9 100.4 95.6 95.1 94.4 92.4 90 86.7 89.6 91.5 91.8 93.1 94.2 96.8 90.5 87.4 88.4 86.2 84.0 86.4
Assessment on spending money on semi-durable goods (next 3 months) 99.1 101.2 104.6 103.1 102 101.7 104.4 106.2 108.8 105.3 104.0 102.0 101.6 102.5 102.1 101.2 102.7 103.8 103.5
9
120.3 125.4 137.9 140.2 141.6 139.8 134.4 139.4 133.3 132.4 132.5 131.8 137.9 132.3 132.8 133.6 133.4 133.1 131.5 134.8 127.1
Buying time condition of durable goods (present)
Table 4 Diffusion Indices based on individual questions concerning the consumer tendency Probability of buying durable goods (next 6 months) 24.5 24.1 26.2 22.5 22.5 23.6 20.4 19.4 19.9 17.5 18.0 20.4 19.2 20.9 20.7 19.4 19.1 19.3 19.4 19.6
Probability of buying a car (next 6 months) 10.1 9.4 9.3 8.8 8.9 9.3 9.8 7.3 9.1 8.4 8.9 7.7 7.9 8.8 8.1 8.0 7.6 7.7 8.6 8.0
Probability of buying or building a home (next 12 months) 9 7.9 8.3 8.6 8.2 8 7.7 6.2 8.6 7.6 7.1 6.8 7.7 9.1 8.3 8.6 8.3 8.0 8.9 8.8
Pobability of spending any money on home improvements or renovations (next 6 months) 32 33.8 35.4 33.8 33.3 33.9 27.7 24.6 23.4 18.8 18.7 18.3 22.2 25.1 28.7 30.6 29.8 27.3 23.7 23.5
Probability of borrowing money to finance consumption expenditures (next 3 months) 45.2 42.4 40.2 38.3 38.3 36.2 36.3 38.1 35.6 37.2 36.9 35.4 35.5 32.7 34.3 33.7 35.2 34.6 34.8
Saving time condition (present) 85.2 92.8 89.2 88.7 90.5 83.6 80.4 79.1 75.5 72.9 72.5 70.7 72.5 73.7 76.2 78.1 73.7 70.0 70.6 69.5 69.8
Probability of saving (next 6 months) 39.1 39.9 32.8 31.5 30.6 30.9 31.3 27.2 25.5 26.4 26.1 24.8 25.6 25.2 24.5 26.2 23.7 22.1 23.1 23.2 21.6
90.1 92.1 88.8 86.4 90.1 80 85.9 82 74.8 74.5 74.3 72.8 77.9 87.3 88.3 87.5 86.0 90.7 87.5 87.9 87.4
Expectation about the direction of price changes (next 12 months)
11. Conclusion In this study the CTS is introduced in a detailed way and the qualitative questions of the CTS are analyzed with the Consumer Confidence Index that has been developed by using the methodology of EU (User Guide, 2003). This study indicates the evaluation of the consumer confidence index's performance, the interpretation of the index and the graphical analysis of the sub-indices with the realizations. The sub-indices of the Consumer Confidence Index are found to be quite reliable and consistent. There are some restrictions in constructing the confidence index. CTS has been conducted since December 2003 therefore the statistical analysis like cross-correlation or seasonality tests could not be evaluated due to insufficient number of observations. It is clear that though there are some limitations of the survey, the Consumer Confidence Index performs quite well. It is worth to note that there is always a possibility to construct different confidence indices based on different criteria. Besides, the explanatory properties of the indices of the Consumer Confidence Index must be revised from time to time based on statistical criteria in order to validate its performance in the future.
10
References User Guide (2003), “The Joint Harmonised EU Programme Of Business And Consumer Surveys”, European Commission Directorate General Economic And Financial Affairs. Curtin, R. (2002), “Surveys of Consumers: Theory, Methods, and Interpretation”, NABE 44th Annual Meeting, Washington DC. Nilsson R. (2000), “Consumer Surveys Methodology, analytical use and presentation of results”, Seminario Sobre Indicadores Lideres y Encuestas de Expectativas, Rio de Janeiro. Kershoff, G. (2000), “Measuring Business and Consumer Confidence in South Africa”, BER, Stellenbosh, December.
11
Appendix Questions of the CTS: 1.
How does your purchasing power at present compared to the past 6 months? Much more better
2.
No idea
A little bit better
Remain the same
A little bit worser
Much more worser
No idea
A little bit better
Remain the same
A little bit worser
Much more worser
No idea
A little bit better
Remain the same
A little bit worser
Much more worser
No idea
How do you think the job opportunities in Turkey will change over the next 6 months? Increase sharply
6.
Much more worser
How do you think the general economic situation in Turkey will change over the next 3 months? Much more better
5.
A little bit worser
How does the current general economic situation in Turkey at present compared to the past 3 months? Much more better
4.
Remain the same
How do you think your own purchasing power will change over the next 6 months? Much more better
3.
A little bit better
Increase slightly
Remain the same
Fall slightly
Fall sharply
No idea
How do you think your spending on semi-durable goods (clothes, shoes, kitchen equipment, etc.) will change over the next 3 months compared with what it was 3 months ago? Increase sharply
7.
Increase slightly
Fall slightly
Fall sharply
No idea
Do you think it is the right time to buy durable consumption goods such as refrigerator, TV, furniture? It is neither the right time nor the wrong time
Yes, it is the right time now
8.
Remain the same
No, it is not the right time now
No idea
How likely are you or your household to buy durable consumption goods like refrigerator, TV, furniture over the next 6 months? Very likely
9.
Fairly likely
Not likely
Not at all likely
No idea
How likely are you or your household to buy a car over the next 6 months? Very likely
Fairly likely
Not likely
Not at all likely
No idea
10. How likely are you or your household to buy or build a home over the next 12 months? Very likely
Fairly likely
Not likely
12
Not at all likely
No idea
11. How likely are you or your household to spend any money on home improvements or renovations (spendings on heating system, whitewash, kitchen-bath repairs, etc. except spendings on little amount of purchases) over the next 6 months? Very likely
Fairly likely
Not likely
Not at all likely
No idea
12. How likely are you or your household to borrow money (consumer credits, other borrowings) to finance consumption expenditures over the next 3 months? Very likely
Fairly likely
Not likely
Not at all likely
No idea
13. In view of the general economic situation, do you think this is: (saving (TL, foreign currency, gold, deposit, financial investment instruments, etc.) time condition) A very good A fairly time to save good time to save
A very bad time to save
Not a good time to save
No idea
14. How likely are you to save (TL, foreign currency, gold, deposit, financial investment instruments, etc.) over the next 6 months? Very likely
Fairly likely
Not likely
Not at all likely
No idea
15. In comparison to the realizations, what do you think about the direction of price changes over the next 12 months? Increase more rapidly
Increase at the same rate
Increase at a slower rate
13
Stay about the same
Fall
No idea
