Successfully reported this slideshow.

Dung PhD Conference 2012

662 views

Published on

  • Be the first to comment

  • Be the first to like this

Dung PhD Conference 2012

  1. 1. Preliminary  results     Dung  Doan   Crawford  School  of  Public  Policy   27  November  2012 Income and diet quality in China
 2004-2009
  2. 2. Research context   Structural shift in dietary consumptions in developing countries   Characterized not by calories deficiency, but by unbalanced diet   Significant impacts on dietary welfare, health, and labor quality   Warnings of deteriorating diet quality as income rises in China   Shift away from Chinese tradition diet   Higher consumption of animal foods and edible oils   Lower grain and vegetable consumption   Calories intake decreases   Increasing income hardly reduces vitamin A & D deficiencies 2
  3. 3. Literature review   Strong focus on diet adequacy   Levels of food and nutrient consumption as dependent variables   E.g. Berhman & Deolalikar (1987), Ravallion (1990), Skoufias (2002, 2009), Mangyo (2008)   Studies on consumption quantity are insufficient to inform about how diet quality changes as income grows   Estimated income elasticity varies widely across studies   Consumption quantity reveals limited info about changes in diet structure and quality   Other aspects of diet quality are under-studied, despite being well-grounded in nutrition literature 3
  4. 4. Research questions   Does diet quality necessarily deteriorate as income rises in China?   Are there income effects and what form do they take?   Do income effects change over time?   Are there education effects?   Contribution:   Directly examine income effects on an essential aspect of diet quality that no economists has paid attention to: diet variety   Most recent time period 4
  5. 5. Data 5   China Health and Nutrition Survey 2004, 2006, 2009   Multi-stage, randomized cluster sampling design   9 provinces, approx. 3,800 households/year   Pool sample: 20,307 adults (18-60 years old) from 4,506 households
  6. 6. Methodology   Dependent variable:   Diet variety, measured by number of major food groups consumed   Repeated cross-sectional regression models   OLS :   Poisson quasi-maximum likelihood: where 6
  7. 7. Income effects 7 OLS   Poisson     Model   Variable   2004   2006   2009   2004   2006   2009   1   Income   0.073***   0.016   0.0092         0.064***   0.016**   0.0094*   2   Income   0.114***   0.088***   0.035***   0.114***   0.087***   0.036***   Income  squared   -­‐0.0082*   -­‐0.0046***   -­‐0.0014***   -­‐0.0094***   -­‐0.0051***   -­‐0.0016***   3   QuinJle  2   0.0074   0.056**   -­‐0.038         0.014   0.076**   -­‐0.038   QuinJle  3   -­‐0.034   0.015   -­‐0.043         -­‐0.024   0.039   -­‐0.041   QuinJle  4   0.108***   0.117***   -­‐0.022         0.119***   0.137***   -­‐0.018   QuinJle  5   0.138***   0.149***   0.101***   0.143***   0.159***   0.096***   *  p  <  0.10,  **  p  <  0.05,  ***  p  <  0.01     Upward concave relationship between income and diet variety   Income effects vary across income groups   Income effects weaken over time
  8. 8. Income effects (cont’) 8   Marginal income effect is positive, stronger among low income groups   Income effects weaken over time 0.00 0.02 0.04 0.06 0.08 0.10 0.12 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Marginal effects at mean income 2004 2006 2009
  9. 9. Education effects   Higher education attainment, higher diet diversity   Difference in impacts of vocational training and university education are not significant at 10% level 9 OLS   Poisson     Educa=on  level   2004   2006   2009   2004   2006   2009   Primary   0.128***   0.032   0.042   0.133   0.036   0.049   Secondary   0.154***   0.106***   0.088***   0.159***   0.107***   0.095***   High  school   0.217***   0.161***   0.150***   0.217***   0.164***   0.158***   VocaJonal  training   0.322***   0.340***   0.296***   0.306***   0.321***   0.287***   University  &  above   0.338***   0.279***   0.376***   0.301***   0.243***   0.341***   *  p  <  0.10,  **  p  <  0.05,  ***  p  <  0.01    Base  educaJon  category  is  “No  educaJon”    
  10. 10. Preliminary conclusion   Household income has positive impact on diet variety.   Income effects are stronger among low income groups, but diminish over time.   Education matters. 10
  11. 11. Issues and Steps forward   Refine model specification and functional form   Address issue of endogeneity between household income and diet variety   Explore a diet quality index as a more comprehensive measure of diet quality 11
  12. 12. Thank you  12
  13. 13. Rationale for diet variety   Diet variety as dependent variable   Unambiguous and well-grounded in nutrition literature   Widely recommended by government’s dietary guidelines   Incorporated in well-known diet quality indexes   Construction of variety measure   Diet variety is most commonly measured by no. of food groups consumed (Ruel 2005)   Food group classification is based Chinese Food Composition Table (Yang et al. 2004, 2009), Chinese dietary guidelines 2007, and Diet quality index – International (Kim et al. 2003) 13
  14. 14. Income effects   Upward concave relationship between income and diet variety   Marginal income effect is smaller among high income groups   Income effects weaken over time – flatter curves Note: The curves represent the predicted quadratic relationship between income and diet variety, while holding all other regressors constant. Thus, values on the vertical axis do not represent nominal value of diet variety. 14 - 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0 3 6 9 12 15 18 Predicteddietvariety Household income per capita (10,000 Yuan) Predicted diet variety based on Model 2 and mean income of percentiles 2004 2006 2009
  15. 15. Income marginal effects   Marginal income effect is smaller among high income groups   Need to refer to nutrition literature to assess the benefit of having 1 additional food group in diet 15 OLS   Poisson     Income  Marginal  effect   2004   2006   2009   2004   2006   2009   Sample  mean   0.099      0.078   0.031     0.097     0.076     0.031     Mean  of  quinJle  1   0.112     0.086     0.034     0.111     0.086     0.035     Mean  of  quinJle  2   0.108      0.084     0.033     0.107     0.083     0.034     Mean  of  quinJle  3   0.104      0.081     0.032     0.102     0.080     0.032     Mean  of  quinJle  4   0.097      0.077      0.030     0.094     0.075     0.030     Mean  of  quinJle  5   0.076      0.061      0.024     0.070     0.058     0.023    
  16. 16. Compare Model 2 and Model 3 16 0.00 0.05 0.10 0.15 0.20 0.25 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Income effects, as compared to Quintile 1 2004 - model 2 2006 - model 2 2009 - model 2 2004 - model 3 2006 - model 3 2009 - model 3
  17. 17. Community-fixed effects OLS   Poisson     Province   2004   2006   2009   2004   2006   2009   Heilongjiang   -­‐0.011   -­‐0.049   -­‐0.088**     -­‐0.0038   -­‐0.0392   -­‐0.076*   Jiangsu   -­‐0.3889**   -­‐0.706***   -­‐0.797***   -­‐0.347***   -­‐0.617***   -­‐0.718***   Shandong   -­‐0.289***   -­‐0.690***   -­‐0.487***   -­‐0.247***   -­‐0.641***   -­‐0.419***   Henan   -­‐0.359***   -­‐0.642***   -­‐0.625***   -­‐0.316***   -­‐0.568***   -­‐0.564***   Hubei   -­‐0.683***   -­‐0.917***   -­‐0.894***   -­‐0.670***   -­‐0.879***   -­‐0.849***   Hunan   -­‐0.453***   -­‐0.800***   -­‐0.881***   -­‐0.434***   -­‐0.722***   -­‐0.829***   Guangxi   -­‐0.838***   -­‐1.216***   -­‐1.175***   -­‐0.991***   -­‐1.369***   -­‐1.088***   Guizhou   -­‐0.435***   -­‐0.6623***   -­‐0.600***   -­‐0.395***   -­‐0.587***   -­‐0.492***   • p  <  0.10,  **  p  <  0.05,  ***  p  <  0.01   • Base  province  is  Liaoning   17   Provincial factors play a significant role in determining diet variety   Considerable differences across some provinces
  18. 18. Checking multicollinearity 18 Model  2   2004   2006   2009   Age  squared   56.37   56.76   55.24   age   56.03   56.45   54.96   HH  income  pc   5.26   4.18   3.74   HH  income  pc  squared   4.53   3.83   3.53   EducaJon  level  -­‐  2   3.21   3.18   3.02   Province_43   2.46   2.74   2.7   EducaJon  level  -­‐  1   2.44   2.58   2.59   EducaJon  level  -­‐  3   2.43   2.51   2.26   Province_32   2.29   2.45   2.13   Province_42   2.25   2.44   2.11   Province_23   2.23   2.41   2.08   Province_37   2.06   2.28   1.92   Province_41   2.04   2.06   1.87   EducaJon  level  -­‐  4   1.84   2.03   1.85   Province_52   1.75   1.94   1.85   Price  -­‐  soy  oil  (log)   1.65   1.91   1.79   EducaJon  level  -­‐  5   1.64   1.88   1.76   Province_45   1.59   1.67   1.54   Price  -­‐  tofu  (log)   1.49   1.62   1.51   Price  -­‐  cabbage  (log)   1.35   1.55   1.42   Rural   1.33   1.41   1.38   Price  -­‐  rice  (log)   1.23   1.38   1.32   Price  -­‐  tofu  (log)   1.21   1.33   1.24   HH  size   1.2   1.27   1.23   Price  -­‐  pork  (log)   1.13   1.24   1.22   Gender   1.06   1.06   1.04   Mean  VIF   6.23   6.31   6.05  
  19. 19. Descriptive Statistics 19 No.  of  food  groups  consumed   2004   2006   2009   Mean   1.72   1.77   1.89   Median   2   2   2   Standard  deviaJon   0.77   0.80   0.83   Percentage   0   0.24%   0.31%   0.10%   1   43.38%   41.76%   36.25%   2   43.30%   41.25%   42.18%   3   10.43%   13.85%   17.94%   4   2.36%   2.56%   3.15%   5   0.29%   0.28%   0.39%   HH  income  pc  (Yuan),  adjusted  to  adult   equivalence  scale   2004   2006   2009   Sample  mean                            9,111                                    10,828                                    14,951     Mean  of  quinJle  1                            1,352                                    1,505                                    1,622     Mean  of  quinJle  2                            3,899                                    4,123                                    6,481     Mean  of  quinJle  3                            6,439                                    7,219                                    10,442     Mean  of  quinJle  4                            10,455                                    11,649                                    16,682     Mean  of  quinJle  5                            23,256                                    28,671                                    38,782    
  20. 20. Dietary guidelines for Chinese residents 20 1.  Eat  a  variety  of  foods,  mainly  cereals  including  appropriate  amount  of  coarse  grains     2.  Consume  plenty  of  vegetables,  fruits  and  tubers       3.  Consume  milk,  soybean  or  dairy-­‐  or  soybean-­‐products  everyday   4.  Consume  appropriate  amounts  of  fish,  poultry,  eggs  and  lean  meat   5.  Use  less  cooking  oil;  choose  a  light  diet  which  is  also  low  in  salt   6.  Do  not  over  eat,  exercise  every  day,  and  maintain  a  healthy  body  weight   7.  RaJonally  distribute  the  daily  food  intake  among  the  three  meals,  correctly  choose  snacks   8.  Drink  sufficient  amount  of  water  every  day,  raJonally  select  beverages     9.  If  you  drink  alcoholic  beverages,  do  so  in  limited  amounts   10.  Choose  fresh  and  sanitary  foods   Source: Ke (2011)
  21. 21. Food classification 21 Food  groups   Foods  sub-­‐groups   Grain   Wheat,  rice,  corn,  barley,  millet,  tubers   Vegetable   Root,  stem,  leafy,  and  flowering  vegetables   Fruit   Fruits   Meat   Pig,  cakle,  mukon,  poultry  ,  fish,  egg,  beans,  and  products   Dairy   Milk  and  dairy  products   Others   Fungi,  nuts  &  seeds,  infant  food,  ethnic  foods  &  cakes,  fast  foods,  beverages,  alcoholic   beverages,  sugars  &  preserves  &  honey,  fats  &  vegetable  oils,  condiments   Source: Chinese Food Composition Tables (Yang et al. 2004, 2009)
  22. 22. Food classification 22 Source: Ke (2011)

×