Some Elements On Consumption Smoothing – A Prima Facie On Household Prediction Of Their Future
I was set on writing a post on social mobility, on the basis of consumption smoothing across time. It was unfortunately a bust, more of a stillborn. But I have the perfect excuse to make for some laziness on my part: data has not been released yet by HCP (for they have promised something early 2011)
Fortunately however, the preliminary computations have not been useless, and provided a certain set of assumptions about household consumption in Morocco, crunched numbers can be of valuable teachings.
The analogy with business cycles is quite important to understand the rest; broadly speaking, the initial, and perhaps the most important, assumption is that households, individuals and businesses, in their rational (or semi-rational) behaviour, tend to favour stability over random and extreme fluctuations. Stability means that everyone appreciates an increase in their consumption of goods, their wealth and their income. In that respect, a head of household would be more content with a steady job and steady prospects too. The same goes for a company in a mature sector, where market shares are very stable, and there is no point in going into a risky endeavour.
Still, an economy like this will die away, because its growth will eventually stop at a certain point. This is why we need to introduce an additional element into the ‘black box’ of the national output: technological progress, a wonderful gift from heaven that translates into better consumption: the best example is that of a housewife with a washing machine (and I can assure the lady readers I am all for shared house-shores) with a completely changed pattern of consumption, with more abundant and better quality goods.
And there lies the first caveat: consumption as depicted in the graph does not account for the increased quality in edible and durable food, the efficiency in electricity consumption, and the rise of the internet, among others. The expressed value merges all of these things into one big aggregate indicator. However, as we are more interested in how consumption as a whole evolves across time, this does not matter a lot. and evolve it did: the average annual increase in consumption per capita has been 0.8%, a 50% increase since 1960 – by comparison, GDP per capita doubled over the same period of time. Even though wealth has considerably increased in Morocco, consumption remained relatively stable: it increased 40% over the last 50 years, but GDP increased almost three times more.
It is also worth noting that GDP and Consumption display a strong correlation in terms of growth rate. indeed, one can even think of some sort of dynamic interaction between value creation and household consumption, a pull/lag of sorts, even more so obvious given the consumption-based GDP aggregate; in this sense, household consumption contributes to growth basis points.
We are however interested in a different topic, namely how well households try to smooth over their consumption across time – regardless of changes in their standards of living. The purpose does have some policy implications: an income tax cut is bound to generate some additional domestic consumption, but by how much and how long is unknown, unless elements such as smoothing consumption across time, with the clear implication that an increase in income not only is conditioned with the immediate marginal propensity of consumption and the smoothing effect for future use of income. Also, there is evidence (later on discussed in detail) on lagging effects -autocorrelation- that go has far as 10 years-time on, perhaps the best evidence that past decisions are factored in over a longer than expected period of time.
The assumption of a perfectly smooth consumption pattern goes right out of the window with the distribution observed, although there is very little to be said about it starting from the 1990s: the trend is steadily decreasing, though it has modestly picked up pace in the late 2000s.
Now, results show that household consumption is a lot less deterministic than I or anyone else would think it to be so. Does it make sense? In a sense, it does. First because households do not necessarily control their outside environment, and exogenous shocks -from productivity and other factors- are more of a random nature than anything else. It is thus understood that past decisions affect present and future ones, and the unaccounted for elements are also lagging over the years.
Truth is, the most persistent pattern starting from “modern” Moroccan economics – modern in the sense of the Adjustment Structural Program aftermath- is a steady decrease in consumption relative to the very long term consumption trend. It is both concerning and interesting: first, it seems as though there is no “trend” to speak of: the trend does exist, but it is only a statistical happening, that there is no collective understanding of the concept. This looks even more alluring that the deviation from the trend is very large, and is very sensitive to brusque historical events: the early 1990s reach a peak because PAS was about to stop, but from then on, there is a steady decrease from the theoretical trend (that coincided with a stiff recession at the time) a very fast deviation from the trend that seems to mimic that of the late 1960s and early 1970s.
There is however a very persistent regularity in these cycles that allow for other means to pursue: there are methods to verify and even predict the next changes in consumption patterns, and these will by the same means produce results on how households will adapt their consumption relative to the trend.
This post is a broad introduction to a more ambitious project seeking a more formal approach to consumption patterns in Morocco – because these are going to be the cornerstone to a host of policies related to taxation, foreign trade, food and edible goods domestic productions, to name but a few.