a guest Feb 28th, 2020 171 Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1. \documentclass{beamer}
  2. \usetheme{Madrid}
  3. \usepackage{amsmath}
  4. \usepackage{natbib}
  5. \bibliographystyle{agsm}
  8. \title{International migration, remittances and household investment: Evidence from Philippine migrants' exchange rate shocks}
  9. \author{\citep{yang2008international}}
  10. \centering
  11. \date{February 2020}
  12. \begin{document}
  13. \maketitle
  14. \begin{frame}{Road-Map}
  15. \begin{itemize}
  16. \item \textbf{Introduction}
  17. \item Literature
  18. \item Theoretical Framework
  19. \item Empirical Strategy
  20. \item Data
  21. \item Results
  22. \item Discussion/Criticisms
  23. \end{itemize}
  24. \end{frame}
  25. \begin{frame}{Introduction}
  26. \framesubtitle{Remittances in Context}
  27. \begin{itemize}
  28.     \item An estimated 2.2-2.9\% of people do not live in their country of origin.
  29.     \item These people often send remittances to their home countries.
  30.     \item Remittances are equivalent to 40\% of FDI inflows in developing countries.
  31.     \item Important but understudied financial flow.
  33. \end{itemize}
  34. \end{frame}
  35. \begin{frame}{Introduction}
  36. \framesubtitle{The Question}
  37. \begin{block}{Three Questions}
  38. The author wants to identify what effect remittances from abroad have on: (1) household consumption, (2) human capital development and (3) entrepreneurial activity.
  39. \end{block}
  40. \end{frame}
  42. \begin{frame}{Introduction}
  43. \framesubtitle{Empirical Challenges}
  44.    \begin{itemize}
  45.        \item \textbf{Problem:} Migrant earnings are not randomly allocated across households.
  46.        \item \textit{Example 1:} More ambitious households send migrants abroad.
  47.        \item \textit{Example 2:} Households that suffer an adverse shock are more likely to send migrants abroad.
  48.        \item \textit{Example 3:} Migrant location is non-random
  49.        \item \textbf{Solution:} Given migrants in various locations, we want to be able to apply random exogenous shocks to the remittances they send home.
  50. \end{itemize}
  51. \end{frame}
  53. \begin{frame}{Introduction}
  54.   \framesubtitle{Set-up}
  55.   \begin{flushleft}
  56.     Exploits the Asian Financial Crisis in 1997 as an exogenous shock to remittances received by Filipino households.
  57.   \end{flushleft}
  58. \begin{block}{Asian Financial Crisis}
  59. Collapse of the Thai Baht caused capital flight in South East Asia. This in turn led to a chain-reaction where many South East Asian currencies depreciated heavily.
  60. \end{block}
  61.   \begin{itemize}
  62.       \item Philippine migrants distributed across various countries.
  63.       \item Size of Philippine peso devaluation different in each country.
  64.       \item Varied largely: e.g US dollar rose in value by 50\% whereas Korean Won fell in value by 4\%.
  65.   \end{itemize}
  66. \end{frame}
  68. \begin{frame}{Introduction}
  69. \framesubtitle{Migrant Locations and Exchange Rate Shock}
  71. \begin{figure}[h!]
  72. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{Selection_146.png}
  73. \end{figure}
  75. \end{frame}
  77. \begin{frame}{Introduction}
  78. \framesubtitle{Summary of Findings}
  79. \begin{flushleft}
  80. Using detailed panel of households before and after the shock, as well as controlling for initial migrant and household characteristics, the author finds:
  81. \end{flushleft}
  82.   \begin{itemize}
  83.       \item \textbf{Human Capital:} Time spent in child labour $\downarrow$. Time child spent in school $\uparrow$. Education expenditure $\uparrow$.
  84.       \item \textbf{Entrepreneurial Activity:} Time devoted to entrepreneurial activity $\uparrow$. Pivot towards capital intensive entrepreneurial activity.
  85.       \item \textbf{Household Consumption: }No significant changes in household consumption.
  86.   \end{itemize}
  87. \end{frame}
  90. \begin{frame}{Road-Map}
  91. \begin{itemize}
  92. \item Introduction
  93. \textbf{\item Literature}
  94. \item Theoretical Framework
  95. \item Empirical Strategy
  96. \item Data
  97. \item Results
  98. \item Discussion/Criticisms
  99. \end{itemize}
  100. \end{frame}
  102. \begin{frame}{Literature}
  103. \framesubtitle{Existing Literature I}
  104. \begin{flushleft}
  105. A largely conflicting literature exists on the effect of remittances.
  106. \end{flushleft}
  107. \begin{block}{Positive Effects}
  108. \begin{itemize}
  109.     \item A host of authors finds that remittances increase savings and investment \citep{brown1994migrants, massey1998international}
  110.     \item Further, authors find that the type of investment associated with remittances is capital-intensive \citep{adams1998remittances, woodruff2007migration}
  114. \end{itemize}
  118. \end{block}
  119. \end{frame}
  121. \begin{frame}{Literature}
  122. \framesubtitle{Existing Literature II}
  123. \begin{block}{Negative Effects/Consumption}
  124. \begin{itemize}
  125.     \item A world bank report from 1991 on the Island of Tonga showed that substantially increased private consumption levels. \citep{world1993pacific}
  126.     \item Studies have also shown that remittances have undermined local development \citep{reichert1981migrant}
  127.     \item Some have expressed concern that remittances can create market distortions and a rentier economy \citep{martin2009labour}
  128.     \item A key example of this distortion examined by \cite{sachs2001curse} is a dutch disease type effect.
  130. \end{itemize}
  131. \end{block}
  132. \end{frame}
  134. \begin{frame}{Literature}
  135. \framesubtitle{Contribution}
  136. \begin{itemize}
  137.     \item Better understand how remittances are spent through the use of random changes in income
  138.     \item Contributes to the literature on transitory shocks
  139.     \item Better understanding of remittances as financial flow
  140. \end{itemize}
  142. \end{frame}
  144. \begin{frame}{Road-Map}
  145. \begin{itemize}
  146. \item Introduction
  147. \item Literature
  148. \item \textbf{Theoretical Framework}
  149. \item Empirical Strategy
  150. \item Data
  151. \item Results
  152. \item Discussion/Criticisms
  153. \end{itemize}
  154. \end{frame}
  156. \begin{frame}{Theoretical Framework}
  157. \begin{itemize}
  158.     \item Standard theory: Transitory shocks to income should have no significant effects given access to credit (consumption smoothing)
  159.     \item In developing countries investments often require upfront fixed costs, individuals are often credit constrained $\implies$ timing of investments is dependent on current income
  160. \end{itemize}
  161. \begin{block}{Case of Child Labour}
  162. Parent has to decide between sending child to school, or utilising them for child labour. In developed country, one could consumption smooth while child is at school, hence no real lost income. In developing country, with no access to credit and high tuition costs for school (upfront cost), this is not viable. Hence if there is a positive shock to income $\implies$ parents can afford to send child to school.
  163. \end{block}
  164. \begin{itemize}
  165.     \item Model predicts that increased remittances $\implies$ investment $\uparrow$ given that there are credit constraints.
  166. \end{itemize}
  167. \end{frame}
  169. \begin{frame}{Road-Map}
  170. \begin{itemize}
  171. \item Introduction
  172. \item Literature
  173. \item Theoretical Framework
  174. \item \textbf{Empirical Strategy}
  175. \item Data
  176. \item Results
  177. \item Discussion/Criticisms
  178. \end{itemize}
  179. \end{frame}
  181. \begin{frame}{Empirical Strategy}
  182. \framesubtitle{Standard Specification I}
  183. \begin{flushleft}
  184. The empirical strategy centres around the following first-differenced specification:
  185. \end{flushleft}
  186. \begin{equation}
  187.     \Delta Y_{it} = \beta_0 + \beta_1(ERSHOCK_i) + \varepsilon_{it}
  188. \end{equation}
  189. \begin{flushleft}Where:
  190. \end{flushleft}
  191. \begin{itemize}
  192.     \item $\Delta Y_{it}$ is the change in outcome variable between 1997 and 1998 (pre and post crisis).
  193.     \item $ERSHOCK_i$ is the exchange rate shock faced by Household \textit{i}, calculated as a weighted average of the exchange rate shocks faced by all migrants in the household.
  194.     \begin{equation}
  195.         ERSHOCK_i = \frac{\sum_{j=1}^J n_{ij}ERCHANGE_j}{\sum_{j=1}^J n_{ij}}
  196.     \end{equation}
  197.     \item $\varepsilon_{it}$ is a mean zero error term.
  198.     % nij is the number of individuals from household i working in country j, so this is just a weighted average of all individuals a family has working abroad.
  199. \end{itemize}
  200. \end{frame}
  202. \begin{frame}{Empirical Strategy}
  203.     \framesubtitle{Standard Specification II}
  204. \begin{flushleft}The regression coefficients then have the following interpretation:
  205. \begin{itemize}
  206.     \item $\beta_0$ accounts for the average change in outcomes across all households in the sample. In this case represents the shared impact of the economic crisis on Philipine households.
  207.     \item $\beta_1$ is the impact of unit change in the exchange rate shock on the outcome variable.
  208. \end{itemize}
  209. Identifying $\beta_1$ as causal requires that the \textbf{common trends assumption} is satisfied.
  210. \begin{itemize}
  211.     \item If households had faced identical exchange rate shocks, changes in outcomes need be the same across households.
  212.     \item Obvious violation would be if migrants in countries with more favourable shocks were characteristically different to those in countries with less favourable shocks, causing outcomes to have evolved differently in the absence of shocks.
  213.     %Talk about how its important that pre characteristics must also affect differential changes in outcomes
  215. \end{itemize}
  216. \end{flushleft}
  217. \end{frame}
  218.  \begin{frame}{Empirical Strategy}
  219.      \framesubtitle{Extended Specification}
  220. \begin{flushleft}
  221. Consider the following example of potential endogeneity:
  222. \begin{itemize}
  223.     \item Post-crisis, small enterprises are less likely to fail in better educated households.
  224.     \item Better educated households may also be sending their migrants to more favourable exchange rate shock locations.
  225. \end{itemize}
  226. Therefore also run a regression controlling for pre-crisis characteristics.
  228. \begin{equation}
  229.     \Delta Y_{it} = \beta_0 + \beta_1(ERSHOCK_i) + \delta '(X_{it-1}) + \varepsilon_{it}
  230. \end{equation}
  231. \begin{itemize}
  232.     \item $X_{it-1}$ is a vector of various geographical and precrisis household/migrant characteristics.
  233.     \item Captures any effect pre-crisis characteristics may have on outcome variables.  
  234. \end{itemize}
  235. \end{flushleft}
  237. \end{frame}
  239. \begin{frame}{Road-Map}
  240. \begin{itemize}
  241. \item Introduction
  242. \item Literature
  243. \item Theoretical Framework
  244. \item Empirical Strategy
  245. \item \textbf{Data}
  246. \item Results
  247. \item Discussion/Criticisms
  248. \end{itemize}
  249. \end{frame}
  251. \begin{frame}{Data}
  252.     \begin{itemize}
  253.         \item Combines data from 4 linked household surveys, conducted by Philippine National Statistics office.
  254.         \item Conducted quarterly, with final survey containing 75\% of the same households surveyed in first period.
  255.         \item Households only included in final sample if they have at least one member working abroad.
  256.     \end{itemize}
  257. \begin{block}{Potential Concern: \textit{Attrition}}
  258. Quite low in the sample at only 5.6\%, but becomes problematic if it is correlated with the ER shock, leading to biased estimators. Author runs a simple test:
  259. \begin{itemize}
  260.     \item Runs regressions where dependent variable takes value 1 if household is trackable throughout surveys and 0 otherwise, and independent variable is the ER shock.
  261.     \item In no regression is coefficient on the ER shock different from zero.
  262. \end{itemize}
  263. \end{block}
  264. %explain why this is problematic in actual presentation
  265. \end{frame}
  267. \begin{frame}{Road-Map}
  268. \begin{itemize}
  269. \item Introduction
  270. \item Literature
  271. \item Theoretical Framework
  272. \item Empirical Strategy
  273. \item Data
  274. \item \textbf{Results}
  275. \item Discussion/Criticisms
  276. \end{itemize}
  277. \end{frame}
  279. \begin{frame}{Results}
  280. \framesubtitle{The Effect of Exchange Rate Shocks on Remittances}
  281. \begin{flushleft}
  282. First, the author shows the relationship between exchange rate shocks and remittances. Column (1) is the standard regression, Column (2) includes controls.
  283. \end{flushleft}
  284. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result1.png}
  286. \end{frame}
  288. \begin{frame}{Results}
  289.     \framesubtitle{The Effect of Remittances on Consumption}
  290. \begin{flushleft}
  291. We see increases in income, results show that these are likely to be due to remittances. No changes in consumption.
  292. \begin{itemize}
  293.     \item One s.d increase ER shock increases total household income by 4.2\%.
  294.     \item No Change in Consumption
  295. \end{itemize}
  296. \end{flushleft}
  297. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result2.png}
  298. \end{frame}
  301. \begin{frame}{Results}
  302. \framesubtitle{The Effect of Remittances on Non Consumption Expenditure}
  303. \begin{flushleft}
  304. Examines the impact of exchange rate shocks on households non-consumption disbursements, expressed as a fraction of initial household consumption. The elasticity is NB!!!
  306. \end{flushleft}
  307. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result3.png}
  308. \end{frame}
  310. \begin{frame}{Results}
  311. \framesubtitle{The Effect of Remittances on Human Capital Development}
  312. \begin{flushleft}
  314. Additionally, the paper finds important results relating to child development indicators. A one s.d increase in size of ER shock:
  315. \begin{itemize}
  316.     \item \textit{Increases} likelihood of being a student by 1.6\%
  317.     \item \textit{Decreases} hours worked by a child in the past week by 0.35.
  318. %There is a small increase in hours worked by boys in certain family related businesses, but is not big enough to offset negative general effects
  319. \end{itemize}
  320. \end{flushleft}
  321. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result4.png}
  322. \end{frame}
  324. \begin{frame}{Results}
  325. \framesubtitle{The Effect of Remittances on Entrepreneurial Activity I}
  326. \begin{flushleft}
  327. We find no changes in hours worked, however we do see a pivot towards entrepreneurial activity. A one standard-deviation increase in the size of the exchange rate shock (0.16) is associated with a differential increase in hours worked in self employment of 1.6 hours per week.
  328. \end{flushleft}
  329. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result5.png}
  330. \end{frame}
  332. \begin{frame}{Results}
  333. \framesubtitle{The Effect of Remittances on Entrepreneurial Activity II}
  334. \begin{flushleft}
  335. We notice that workers are moving into new activities. A one standard-deviation increase in the size of the exchange rate shock (0.16) is associated with a differential increase in the likelihood of entering a new entrepreneurial activity of 2.2 percentage points.
  336. \end{flushleft}
  337. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result6.png}
  338. \begin{flushleft}
  339. Yang only observes statistically significant movements into entrepreneurial activities with high fixed costs. Following one s.d increase in ER shock:
  340. \begin{itemize}
  341.     \item Transportation/communications and manufacturing rise by 1.2\% and 0.9\% respectively. Very large as amount of households in these areas before was low.
  342.     \item Tentative evidence of a decline in farming/gardening and wholesale retail/trade - indicates a possible shifting away to better activities.
  343. \end{itemize}
  344. %TEXT HERE
  346. \end{flushleft}
  347. \end{frame}
  349. \begin{frame}{Results}
  350. \framesubtitle{The Effect of Remittances on Entrepreneurial Activity III}
  351. \begin{flushleft}
  353. \end{flushleft}
  354. \includegraphics[width=\textwidth,height=\textheight,keepaspectratio]{result7.png}
  355. \end{frame}
  357. \begin{frame}{Results}
  358.     \framesubtitle{Robustness}
  359.     \begin{flushleft}
  360.     Are these effects capturing the exchange rate shock or other economic shocks associated with the financial crisis?
  361.     \begin{itemize}
  362.         \item Attempts to control for real economic shocks.
  363.         \item Include a \textit{job loss indicator} to control for whether migrant lost their job overseas.
  364.         \item Further, include a control for \textit{change in log GDP} in migrant destination countries, between 1996 and 1998.
  366.     \end{itemize}
  367. \textit{\textbf{Result:}} Inclusion has very little effect of coefficients of interest.
  368. \begin{itemize}
  369.     \item E.g entrepreneurial entry coefficient from 0.139 to 0.128; schooling coefficient from 0.091 to 0.093.
  370.     \item Exchange rate shocks themselves likely to be primary causal factor.
  371. \end{itemize}
  372.     \end{flushleft}
  373. \end{frame}
  375. \begin{frame}{Road-Map}
  376. \begin{itemize}
  377. \item Introduction
  378. \item Literature
  379. \item Theoretical Framework
  380. \item Empirical Strategy
  381. \item Data
  382. \item Results
  383. \item \textbf{Discussion/Criticisms}
  384. \end{itemize}
  385. \end{frame}
  387. \begin{frame}{Discussion}
  388. \begin{itemize}
  389.     \item Paper ultimately showed that remittances positive in the normative sense.
  390.     \item Vast policy implications for this.
  391. \end{itemize}
  392. \begin{block}{Policy Example}
  393. Consider pro-migrant policy like past proposals to legalise the undocumented migrants in the US. It might be easier for them to find work and send remittances back home. This will have large positive effects for home countries. Can reduce migration in the long run.
  394. \end{block}
  395. \begin{flushleft}
  396. However, we have some criticisms:
  397. \end{flushleft}
  398. \begin{itemize}
  400.     \item \textit{Critique 1:} Paper is not necessarily externally valid. A financial crisis is not a normal event.
  401.     \item \textit{Critique 2:} No control group.
  402.     \item \textit{Critique 3:} No examination of general equilibrium effects. Also, not possible to examine certain GE effects (such as dutch disease) due to financial crisis.
  403. \end{itemize}
  405. \end{frame}
  407. \begin{frame}[allowframebreaks]{References}
  408. \bibliography{ref}
  409. \end{frame}
  411. \begin{frame}
  412. \includegraphics[width=60mm,keepaspectratio]{shaun thumbs.jpg}
  413. \huge{\centerline{Thanks for listening! Questions?}}
  414. \end{frame}
  417. \end{document}
RAW Paste Data
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy. OK, I Understand