I would like to make a distinction.
Rising prices is not synonomous with inflation. For example the price of tickets to an AC/DC concert has gone up 100 fold since the 1970's, not because of inflation but because there is an increased demand for AC/DC tickets. We would not classify a price change in one good as inflation. Leon Walras who is perhaps the forefather of modern economics, came with a nifty theorem known as Walras law. Essentially the idea is that if people spend more on one good they have to spend less on another. This is intuitive as presumedly the 500 000 people who spent $150 to buy AC/DC tickets the other week have $150 less to spend on other items within the economy. So this idea of 'bidding prices up' can work within a partial equilibrium situation, but it is balanced out somewhere else within the economy. This is also true of situations where people work less and subsequently earn lower incomes. This can simply be viewed as people purchasing more leisure time and as a result they have to spend less on other areas within the economy. This provides a bit of a measurement conundrum for policy makers as it is hard to include items like 'leisure' in the CPI. Nonetheless the key point I would like to make is that changing prices is not the same thing as inflation.
So this leads to the very interesting question of what is inflation. Well in order to do this we need to define inflation. Of course you already know this as 'a sustained increase in prices'. But what do we mean by prices? Well in a modern monetary economy it refers to the amount of notes or coins you need to obtain a good or a service. But historically this is not how it always happened. Consider a two good exchange economy without a monetary system. The price of good 1 is essentially the amount of good 2 that you would need in order to obtain good 1. Now a legitimate change in price would occur if you need more of good 2 in order to obtain good 1 than you did in some previous period. This is not inflation, it is just a price change. Inflation in such a system is hard to imagine, but I suppose it can be demonstrated by the following example:
1 kg of rice = 2 kg of wheat (Year 1)
2 kg of rice = 4 kg of wheat (Year 2)
In this case the exchange relationship is the same its just that you need more of one thing to buy more of another (in the same ratio). Now as I alluded to, inflation is hard to imagine as in this example you can simply say that the exchange ratio is the same for years 1 and 2. This however is exactly what happens when we introduce inflation. Consider the same economy with the introduction of money by a central bank. The way to consider the introduction of money is as follows. Consider it as a third good, whose value is totally derived from the value of another asset. I.e. it holds no intrinsic value, but rather an exchange value. Because its value is derived only from its exchange value, money could be anything. It could be notes or coins, seashells, gold, or even cow dung. Now consider the following example:
1kg of rice = $100
1kg of wheat = $50 (Year 1)
1kg of rice = $200
1kg of wheat = $100 (Year 2)
Notice how this example corresponds to the previous one? we have maintained the exchange ratio's, but doubled the price. This is equivalent to the first example. This is essentially inflation. Unlike the case of a simple price change, prices have risen without any change in the price ratio. This in the purest of terms is what is meant by inflation. The introduction of money makes life easier for economists as it allows them to introduce a 'numeraire'. You can imagine how hard it would be to model the economy if the price of one good was represented in all the exchange values of other goods. Indeed the advantage of money is better shown in its application within the real economy. It acts first and foremost as an information tool that demonstrates to a consumer the relative exchange relationships between goods. We still essentially have a barter society. You produce your labour for money which is one form of barter, and then you exchange your money for another good or service which is another kind of barter.
So where does inflation come from? well this is really interesting. I am going to restrict my analysis to the classical approach but I add a word of warning by saying that in reality inflation is far more complex. One of the things that is essential for money or the 'numeraire' to be effective, is its scarcity. For this reason gold and silver was a popular numeraire in earlier times. Indeed when gold was being traded as 'money' it was traded independent of any kind of formal authority from the government or a central bank. Early consumers kind of figured out that it would be easier to trade in gold rather than doing a direct exchange. Scarcity is important because if the supply of money increases, then you have more money chasing the same amount of goods which results in inflation. An instructive example comes from a recent television program I watched. It showed how the spanish mined gold in south america and as a result they had an increase in the money supply. But the only impact that this had was an increase in prices. So this leads me to my next point. Where does modern inflation come from? Well if you follow the classical line of thought, inflation comes from an increase in the money supply. Modern central banks issue currency that can be thought of as being scarce. People have faith that the currency will remain scarce, and subsequently are confident in holding money and using it for exchange. Traditionally central banks have ensured the scarcity of money, and built public confidence by backing currency with gold. The gold standard was required in order to convince people that they could use the notes and coins issued by the Bank of England (who was the first to adopt such a system). However in modern economies we no longer have a gold standard and have instead moved toward 'fiat' money which is essentially where the money has no inherent value (as it is not backed by gold or some other kind of underlying asset). The reason fiat money works is because people trust that central banks will not excessively expand the money supply and that it is widely used as a medium of exchange through which it derives its value.
So how does this relate to inflation? Well this is an important introduction into some of the issues in creating new money. I.e. when, where, and why should the central bank expand the money supply. In economics the equation that was championed by the classical economists and the monetarists was the velocity of money equation. I.e.:
MV = PY
where M equals the amount of money within the economy, V equals the velocity of money which is the amount of times each unit of money will be used, P which is the price level, and Y is production. Essentially this is saying that the money stock multiplied by the amount of times it changes hands will equal the total value of production (i.e. PY). This intuitively makes sense because if there are ten dollars within the economy, and 100 goods which are priced at one dollar each, the ten dollars within the economy will have to be spent 10 times in order for the money supply to facilitate this expenditure. I.e. M(10) x V(10) = P(1) x Y(100). So this leads to an interesting observation. What happens if we increase the money supply (M) whilst keeping everything else constant? Well one of two things has to happen. Either the price increases, or the amount of goods and services produced will increase. Now in the long run production will be unaffected so it must impact on the price level. In the short run this may produce what is referred to as a 'money illusion' effect whereby people perceive that their incomes are higher and subsequently spend more (which may account for an increase in Y). So in a classical sense this is where inflation comes from - excessive growth in the money supply. This explains why nations like Zimbabwe have hyperinflation. Zimbabwe has used 'seigniorage' in order to raise government revenues. This is just a nice word for 'printing money'. The massive increases in the money supply has resulted in massive inflation and lost confidence in the currency.
Milton Friedman in the 1960's and 70's spearheaded a movement known as monetarism which re-emphasised these relationships. whilst monetarism ultimately failed as its policy implications proved impractical to implement; it certainly generated a shift towards low inflation policies. This is why central banks tend to target inflation.
To comment on Scora's comments and the question more specifically, 0% inflation would not be anything particularly remarkable. We have achieved 0% inflation before and it was nothing out of the ordinary. Indeed many countries (e.g. Japan) have had significant levels of deflation which has its own unique challenges. 0% inflation in the purest of senses would be optimal however there are a number of reasons why a small low positive level of inflation can be preferrable given the realities of the economy. Firstly inflation measurement issues can mean that there is a positive inflation bias. I.e. the quality of products are not considered so prices may increase but that could be due to a quality improvement. Secondly deflation can create liquidity trap risks which can see central banks prefer positive inflation. Thirdly the labour market is sticky downwards so inflation can provide a way for real wages to fall which may assist the labour market in clearing. There are a whole heap, but the traditional phillips curve trade off is not usually thought to be one of them. This in part is due to the long run phillips curve and the objections of monetarism.
