TFP is generally estimated as a residual – the difference between a firm's actual output and its predicted output. This requires a two-step procedure: first we predict the value of output based on information on a firm's use of inputs, a theoretical framework explaining how the inputs are combined and a consistent empirical estimation that measures this combination. Then we take the difference between the (correctly estimated) predicted output and the actual (observed) output. This residual will explain which part of the actual output we are not able to explain by looking at the way that firms have used inputs.

We have focused on labour allocation as another factor that helps explain aggregate economic growth through its effects on productivity. The effect of labour misallocation is potentially large in all countries for which we have data. Any policy fostering reallocation of workers needs jobs to churn in the first place. Only if workers lose jobs in unproductive firms can they find alternative occupation in more productive firms. Policies promoting labour reallocation should be accompanied by active labour market policies to reduce the displacement costs of workers, or specific policies aimed at reducing potential wage disparity. Increasing the minimum wage is thus a policy which could have positive reallocation effects for the economy. But if small firms are forced out of business as a result, the same policy might also reduce competition and lead to less choice for consumers.

Using “average" indicators is simply misleading in most occasions, and firm-level analysis allows a much more targeted set of policies. Firm-level analysis is an essential tool to complement and integrate the macro assessment and related policy response across the whole range of productivity drivers – from labour to trade, from finance to competition. The great advantage of the CompNet dataset is that is built in such a fashion to be able to be used – as it is and directly – to derive at the very least a first set of granular stylised facts to inform policy considerations. Users can use the dataset to assess competitiveness, financial constraints and sensitivity to exchange rate fluctuations. The book has also presented a wide variety of applications of firm-level-based analysis which goes over and above the mere presentation of stylised facts and joint distributions coming directly out of the dataset, including zombie firms impacts, export and participation in global value chains and concentration and market power.

Until recently, data limitations made it almost impossible to conduct microeconomic cross-country research. In this chapter, we present the most relevant micro-level datasets, which are available and are being used both in academia and in policy institutions, including the EU-Klems project, the World Input Output Database (WIOD), Orbis and its European version Amadeus, DynEmp, MultiProd, EFIGE and, finally, the Competitiveness Research Network (CompNet).

Productivity and growth are affected by the financial environment as a whole. This environment is a combination of firms' financial conditions (the demand side), and a country's quality of financing institutions (the supply side). A malfunctioning financial environment reduces the chance that growing companies can access the external financial resources they need to undertake productive investment. Stagnating productivity growth follows, creating low aggregate economic growth. Firms in countries that underwent serious financial distress due to the financial crisis have persistently suffered from its consequences. The situation led to a piling up of debts, a contraction of investments and a strong decrease in firms' profitability. This process also led (and is partially due) to an increase in the number of low-productivity zombie firms in periphery countries after the financial crisis.

Productivity diverges widely and persistently between – and most notably within – countries and industries. Connecting the micro and macro perspective is also a challenge productivity research has to face. To tackle such issues, since the early 2010s the Competitiveness Research Network – CompNet – a large European initiative, was created with two main objectives: (i) developing a forum for discussion on productivity research; and (ii) setting up a truly comparable EU dataset of productivity indicators, based on firm-level information of top quality. We will extensively use the research work and the dataset generated by CompNet in this book. Our ultimate objective is to show how we can gather rich and critical information by taking a firm-level perspective, but also argue that despite the progress made, additional EU-wide efforts need to be made on the data front to ensure suffcient depth and accurateness.

Firm heterogeneity is nowadays at the center of trade analysis. There seems to be a link between firms' presence on foreign markets and productivity of individual companies. There is a premium for companies engaging in internationalization activities against those serving only the local market, and these premia seem to increase with the complexity of the internationalization activity. There is causality between firms' innovation and internationalization activities. This implies that innovation policies are a major driver of trade and economic growth. Few firms are able to export, and export is thus very concentrated in a few firms in virtually all countries. This fact means that policymakers might be able to analyse and forecast nearly all aggregate trade figures by predicting a few firms' export value. Responses to exchange rate movements are heterogeneous across firms: export elasticity to REER movements is inversely correlated with size and productivity, and therefore exports by larger and more productive companies are less sensitive to exchange rate movements.

For those countries or industries in which a rise in (measured) market power actually translates into an effective restriction of competition and increasing profits, as it seems to be the case for the US, a significant slowdown of investment, business dynamism, fall in labour share and wages has been observed. The latter in turn has several policy implications, from obvious considerations of reduced consumers' welfare, to low productivity growth, to severe implications for the transmission mechanism of monetary policy. However, one has to question the nature of such an increase in market power. If the latter stems from “good” market forces in which industries experience a reallocation of economic activities towards the largest and most productive firms, as some research shows, a debate opens up on the necessary reform of competition policies.

It is necessary to distinguish the productivity of a single production factor from the ability of a firm to combine the bundle of inputs at its disposal to create new value. This is total factor productivity (TFP), also known as multi-factor productivity". Firm performance in a sector or region is not centred around an average performance level. Policies should aim to catalyse the dynamic transition of more-productive firms, rather than just working on the average performance of the sector.

We present in this chapter the empirical techniques used to measure market efficiency and its implications, starting with the so-called OP gap and its dynamic version. We will then discuss the Foster productivity decomposition method and the Hsieh and Klenow techniques to measure allocative inefficiency. Market inefficiency is also related to concentration and market power. When competition in a market is reduced, aggregate productivity growth may decrease, reducing consumer welfare.